diff --git "a/checkpoint-39500/trainer_state.json" "b/checkpoint-39500/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-39500/trainer_state.json" @@ -0,0 +1,276533 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 31.549520766773163, + "eval_steps": 500, + "global_step": 39500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0007987220447284345, + "grad_norm": 3.421875, + "learning_rate": 0.0001, + "loss": 3.5585, + "step": 1 + }, + { + "epoch": 0.001597444089456869, + "grad_norm": 8.125, + "learning_rate": 0.0001, + "loss": 3.5557, + "step": 2 + }, + { + "epoch": 0.0023961661341853034, + "grad_norm": 3.53125, + "learning_rate": 0.0001, + "loss": 2.8222, + "step": 3 + }, + { + "epoch": 0.003194888178913738, + "grad_norm": 2.53125, + "learning_rate": 0.0001, + "loss": 2.4124, + "step": 4 + }, + { + "epoch": 0.003993610223642172, + "grad_norm": 1.890625, + "learning_rate": 0.0001, + "loss": 2.1393, + "step": 5 + }, + { + "epoch": 0.004792332268370607, + "grad_norm": 2.0625, + "learning_rate": 0.0001, + "loss": 1.9346, + "step": 6 + }, + { + "epoch": 0.005591054313099041, + "grad_norm": 1.2578125, + "learning_rate": 0.0001, + "loss": 1.773, + "step": 7 + }, + { + "epoch": 0.006389776357827476, + "grad_norm": 0.90625, + "learning_rate": 0.0001, + "loss": 1.6694, + "step": 8 + }, + { + "epoch": 0.00718849840255591, + "grad_norm": 0.734375, + "learning_rate": 0.0001, + "loss": 1.6065, + "step": 9 + }, + { + "epoch": 0.007987220447284345, + "grad_norm": 0.71484375, + "learning_rate": 0.0001, + "loss": 1.5557, + "step": 10 + }, + { + "epoch": 0.00878594249201278, + "grad_norm": 0.5625, + "learning_rate": 0.0001, + "loss": 1.5308, + "step": 11 + }, + { + "epoch": 0.009584664536741214, + "grad_norm": 0.51953125, + "learning_rate": 0.0001, + "loss": 1.4988, + "step": 12 + }, + { + "epoch": 0.010383386581469648, + "grad_norm": 0.515625, + "learning_rate": 0.0001, + "loss": 1.4728, + "step": 13 + }, + { + "epoch": 0.011182108626198083, + "grad_norm": 0.4296875, + "learning_rate": 0.0001, + "loss": 1.465, + "step": 14 + }, + { + "epoch": 0.011980830670926517, + "grad_norm": 0.3984375, + "learning_rate": 0.0001, + "loss": 1.4369, + "step": 15 + }, + { + "epoch": 0.012779552715654952, + "grad_norm": 0.365234375, + "learning_rate": 0.0001, + "loss": 1.4184, + "step": 16 + }, + { + "epoch": 0.013578274760383386, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.4163, + "step": 17 + }, + { + "epoch": 0.01437699680511182, + "grad_norm": 0.353515625, + "learning_rate": 0.0001, + "loss": 1.4047, + "step": 18 + }, + { + "epoch": 0.015175718849840255, + "grad_norm": 0.328125, + "learning_rate": 0.0001, + "loss": 1.3898, + "step": 19 + }, + { + "epoch": 0.01597444089456869, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.3742, + "step": 20 + }, + { + "epoch": 0.016773162939297124, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.3732, + "step": 21 + }, + { + "epoch": 0.01757188498402556, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.369, + "step": 22 + }, + { + "epoch": 0.018370607028753993, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.3731, + "step": 23 + }, + { + "epoch": 0.019169329073482427, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.356, + "step": 24 + }, + { + "epoch": 0.019968051118210862, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.3473, + "step": 25 + }, + { + "epoch": 0.020766773162939296, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.353, + "step": 26 + }, + { + "epoch": 0.02156549520766773, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.3414, + "step": 27 + }, + { + "epoch": 0.022364217252396165, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.3415, + "step": 28 + }, + { + "epoch": 0.0231629392971246, + "grad_norm": 0.55078125, + "learning_rate": 0.0001, + "loss": 1.3466, + "step": 29 + }, + { + "epoch": 0.023961661341853034, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.3361, + "step": 30 + }, + { + "epoch": 0.02476038338658147, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.3362, + "step": 31 + }, + { + "epoch": 0.025559105431309903, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.3376, + "step": 32 + }, + { + "epoch": 0.026357827476038338, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.334, + "step": 33 + }, + { + "epoch": 0.027156549520766772, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.3228, + "step": 34 + }, + { + "epoch": 0.027955271565495207, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.3177, + "step": 35 + }, + { + "epoch": 0.02875399361022364, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.3312, + "step": 36 + }, + { + "epoch": 0.029552715654952075, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.3358, + "step": 37 + }, + { + "epoch": 0.03035143769968051, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.3224, + "step": 38 + }, + { + "epoch": 0.031150159744408944, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.3086, + "step": 39 + }, + { + "epoch": 0.03194888178913738, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.311, + "step": 40 + }, + { + "epoch": 0.03274760383386582, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.3127, + "step": 41 + }, + { + "epoch": 0.03354632587859425, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.3074, + "step": 42 + }, + { + "epoch": 0.034345047923322686, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.3172, + "step": 43 + }, + { + "epoch": 0.03514376996805112, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.3139, + "step": 44 + }, + { + "epoch": 0.035942492012779555, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.2993, + "step": 45 + }, + { + "epoch": 0.036741214057507986, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.3002, + "step": 46 + }, + { + "epoch": 0.037539936102236424, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.2972, + "step": 47 + }, + { + "epoch": 0.038338658146964855, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.3078, + "step": 48 + }, + { + "epoch": 0.03913738019169329, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.3073, + "step": 49 + }, + { + "epoch": 0.039936102236421724, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.2896, + "step": 50 + }, + { + "epoch": 0.04073482428115016, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.2998, + "step": 51 + }, + { + "epoch": 0.04153354632587859, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.2985, + "step": 52 + }, + { + "epoch": 0.04233226837060703, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.2984, + "step": 53 + }, + { + "epoch": 0.04313099041533546, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.3124, + "step": 54 + }, + { + "epoch": 0.0439297124600639, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.3061, + "step": 55 + }, + { + "epoch": 0.04472843450479233, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.2909, + "step": 56 + }, + { + "epoch": 0.04552715654952077, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.2885, + "step": 57 + }, + { + "epoch": 0.0463258785942492, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.2982, + "step": 58 + }, + { + "epoch": 0.04712460063897764, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.3052, + "step": 59 + }, + { + "epoch": 0.04792332268370607, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.2851, + "step": 60 + }, + { + "epoch": 0.048722044728434506, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.2941, + "step": 61 + }, + { + "epoch": 0.04952076677316294, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.2896, + "step": 62 + }, + { + "epoch": 0.050319488817891375, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.2822, + "step": 63 + }, + { + "epoch": 0.051118210862619806, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.2843, + "step": 64 + }, + { + "epoch": 0.051916932907348244, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.2868, + "step": 65 + }, + { + "epoch": 0.052715654952076675, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.2796, + "step": 66 + }, + { + "epoch": 0.05351437699680511, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.2854, + "step": 67 + }, + { + "epoch": 0.054313099041533544, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.2773, + "step": 68 + }, + { + "epoch": 0.05511182108626198, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.2848, + "step": 69 + }, + { + "epoch": 0.05591054313099041, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.2967, + "step": 70 + }, + { + "epoch": 0.05670926517571885, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.2793, + "step": 71 + }, + { + "epoch": 0.05750798722044728, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.2847, + "step": 72 + }, + { + "epoch": 0.05830670926517572, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.2844, + "step": 73 + }, + { + "epoch": 0.05910543130990415, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.2884, + "step": 74 + }, + { + "epoch": 0.05990415335463259, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.2865, + "step": 75 + }, + { + "epoch": 0.06070287539936102, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.2869, + "step": 76 + }, + { + "epoch": 0.06150159744408946, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.2727, + "step": 77 + }, + { + "epoch": 0.06230031948881789, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.2695, + "step": 78 + }, + { + "epoch": 0.06309904153354633, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.2824, + "step": 79 + }, + { + "epoch": 0.06389776357827476, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.3009, + "step": 80 + }, + { + "epoch": 0.06469648562300319, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.2908, + "step": 81 + }, + { + "epoch": 0.06549520766773163, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.2794, + "step": 82 + }, + { + "epoch": 0.06629392971246006, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.274, + "step": 83 + }, + { + "epoch": 0.0670926517571885, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.2756, + "step": 84 + }, + { + "epoch": 0.06789137380191693, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.2762, + "step": 85 + }, + { + "epoch": 0.06869009584664537, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.2614, + "step": 86 + }, + { + "epoch": 0.0694888178913738, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.2833, + "step": 87 + }, + { + "epoch": 0.07028753993610223, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.2802, + "step": 88 + }, + { + "epoch": 0.07108626198083066, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.2835, + "step": 89 + }, + { + "epoch": 0.07188498402555911, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.2747, + "step": 90 + }, + { + "epoch": 0.07268370607028754, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.2771, + "step": 91 + }, + { + "epoch": 0.07348242811501597, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.2786, + "step": 92 + }, + { + "epoch": 0.0742811501597444, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.2688, + "step": 93 + }, + { + "epoch": 0.07507987220447285, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.2812, + "step": 94 + }, + { + "epoch": 0.07587859424920128, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.274, + "step": 95 + }, + { + "epoch": 0.07667731629392971, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.2663, + "step": 96 + }, + { + "epoch": 0.07747603833865814, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.2712, + "step": 97 + }, + { + "epoch": 0.07827476038338659, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.27, + "step": 98 + }, + { + "epoch": 0.07907348242811502, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.261, + "step": 99 + }, + { + "epoch": 0.07987220447284345, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.2675, + "step": 100 + }, + { + "epoch": 0.08067092651757188, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.2693, + "step": 101 + }, + { + "epoch": 0.08146964856230032, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.2673, + "step": 102 + }, + { + "epoch": 0.08226837060702875, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.2512, + "step": 103 + }, + { + "epoch": 0.08306709265175719, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.2792, + "step": 104 + }, + { + "epoch": 0.08386581469648563, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.266, + "step": 105 + }, + { + "epoch": 0.08466453674121406, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.276, + "step": 106 + }, + { + "epoch": 0.08546325878594249, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.2625, + "step": 107 + }, + { + "epoch": 0.08626198083067092, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.2664, + "step": 108 + }, + { + "epoch": 0.08706070287539937, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.2692, + "step": 109 + }, + { + "epoch": 0.0878594249201278, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.2798, + "step": 110 + }, + { + "epoch": 0.08865814696485623, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.2637, + "step": 111 + }, + { + "epoch": 0.08945686900958466, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.2509, + "step": 112 + }, + { + "epoch": 0.0902555910543131, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.2632, + "step": 113 + }, + { + "epoch": 0.09105431309904154, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.2498, + "step": 114 + }, + { + "epoch": 0.09185303514376997, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.2664, + "step": 115 + }, + { + "epoch": 0.0926517571884984, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.2533, + "step": 116 + }, + { + "epoch": 0.09345047923322684, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.2566, + "step": 117 + }, + { + "epoch": 0.09424920127795527, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.2543, + "step": 118 + }, + { + "epoch": 0.0950479233226837, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.2591, + "step": 119 + }, + { + "epoch": 0.09584664536741214, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.2687, + "step": 120 + }, + { + "epoch": 0.09664536741214058, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.264, + "step": 121 + }, + { + "epoch": 0.09744408945686901, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.2571, + "step": 122 + }, + { + "epoch": 0.09824281150159744, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.2524, + "step": 123 + }, + { + "epoch": 0.09904153354632587, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.252, + "step": 124 + }, + { + "epoch": 0.09984025559105432, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.2444, + "step": 125 + }, + { + "epoch": 0.10063897763578275, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.2539, + "step": 126 + }, + { + "epoch": 0.10143769968051118, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.2598, + "step": 127 + }, + { + "epoch": 0.10223642172523961, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.248, + "step": 128 + }, + { + "epoch": 0.10303514376996806, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.2581, + "step": 129 + }, + { + "epoch": 0.10383386581469649, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.261, + "step": 130 + }, + { + "epoch": 0.10463258785942492, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.251, + "step": 131 + }, + { + "epoch": 0.10543130990415335, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.2579, + "step": 132 + }, + { + "epoch": 0.1062300319488818, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.2509, + "step": 133 + }, + { + "epoch": 0.10702875399361023, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.2483, + "step": 134 + }, + { + "epoch": 0.10782747603833866, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.2482, + "step": 135 + }, + { + "epoch": 0.10862619808306709, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.2426, + "step": 136 + }, + { + "epoch": 0.10942492012779553, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.2524, + "step": 137 + }, + { + "epoch": 0.11022364217252396, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.2491, + "step": 138 + }, + { + "epoch": 0.1110223642172524, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.2428, + "step": 139 + }, + { + "epoch": 0.11182108626198083, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.2571, + "step": 140 + }, + { + "epoch": 0.11261980830670927, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.2475, + "step": 141 + }, + { + "epoch": 0.1134185303514377, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.2535, + "step": 142 + }, + { + "epoch": 0.11421725239616613, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.2538, + "step": 143 + }, + { + "epoch": 0.11501597444089456, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.254, + "step": 144 + }, + { + "epoch": 0.11581469648562301, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.2476, + "step": 145 + }, + { + "epoch": 0.11661341853035144, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.255, + "step": 146 + }, + { + "epoch": 0.11741214057507987, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.2545, + "step": 147 + }, + { + "epoch": 0.1182108626198083, + "grad_norm": 0.37109375, + "learning_rate": 0.0001, + "loss": 1.252, + "step": 148 + }, + { + "epoch": 0.11900958466453675, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.2464, + "step": 149 + }, + { + "epoch": 0.11980830670926518, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.2446, + "step": 150 + }, + { + "epoch": 0.12060702875399361, + "grad_norm": 0.4375, + "learning_rate": 0.0001, + "loss": 1.2517, + "step": 151 + }, + { + "epoch": 0.12140575079872204, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.2414, + "step": 152 + }, + { + "epoch": 0.12220447284345048, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.2508, + "step": 153 + }, + { + "epoch": 0.12300319488817892, + "grad_norm": 0.4140625, + "learning_rate": 0.0001, + "loss": 1.2351, + "step": 154 + }, + { + "epoch": 0.12380191693290735, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.25, + "step": 155 + }, + { + "epoch": 0.12460063897763578, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.249, + "step": 156 + }, + { + "epoch": 0.1253993610223642, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.2395, + "step": 157 + }, + { + "epoch": 0.12619808306709265, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.2549, + "step": 158 + }, + { + "epoch": 0.1269968051118211, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.2344, + "step": 159 + }, + { + "epoch": 0.12779552715654952, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.241, + "step": 160 + }, + { + "epoch": 0.12859424920127796, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.2383, + "step": 161 + }, + { + "epoch": 0.12939297124600638, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.2383, + "step": 162 + }, + { + "epoch": 0.13019169329073482, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.2436, + "step": 163 + }, + { + "epoch": 0.13099041533546327, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.2424, + "step": 164 + }, + { + "epoch": 0.13178913738019168, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.2526, + "step": 165 + }, + { + "epoch": 0.13258785942492013, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.2238, + "step": 166 + }, + { + "epoch": 0.13338658146964857, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.2406, + "step": 167 + }, + { + "epoch": 0.134185303514377, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.2413, + "step": 168 + }, + { + "epoch": 0.13498402555910544, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.2372, + "step": 169 + }, + { + "epoch": 0.13578274760383385, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.2291, + "step": 170 + }, + { + "epoch": 0.1365814696485623, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.2322, + "step": 171 + }, + { + "epoch": 0.13738019169329074, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.2315, + "step": 172 + }, + { + "epoch": 0.13817891373801916, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.2361, + "step": 173 + }, + { + "epoch": 0.1389776357827476, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.2269, + "step": 174 + }, + { + "epoch": 0.13977635782747605, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.2369, + "step": 175 + }, + { + "epoch": 0.14057507987220447, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.2406, + "step": 176 + }, + { + "epoch": 0.1413738019169329, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.2223, + "step": 177 + }, + { + "epoch": 0.14217252396166133, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.2349, + "step": 178 + }, + { + "epoch": 0.14297124600638977, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.2422, + "step": 179 + }, + { + "epoch": 0.14376996805111822, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.2247, + "step": 180 + }, + { + "epoch": 0.14456869009584664, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.2401, + "step": 181 + }, + { + "epoch": 0.14536741214057508, + "grad_norm": 0.390625, + "learning_rate": 0.0001, + "loss": 1.226, + "step": 182 + }, + { + "epoch": 0.14616613418530353, + "grad_norm": 0.640625, + "learning_rate": 0.0001, + "loss": 1.2292, + "step": 183 + }, + { + "epoch": 0.14696485623003194, + "grad_norm": 0.84375, + "learning_rate": 0.0001, + "loss": 1.2437, + "step": 184 + }, + { + "epoch": 0.1477635782747604, + "grad_norm": 1.078125, + "learning_rate": 0.0001, + "loss": 1.2565, + "step": 185 + }, + { + "epoch": 0.1485623003194888, + "grad_norm": 0.470703125, + "learning_rate": 0.0001, + "loss": 1.2407, + "step": 186 + }, + { + "epoch": 0.14936102236421725, + "grad_norm": 0.455078125, + "learning_rate": 0.0001, + "loss": 1.2234, + "step": 187 + }, + { + "epoch": 0.1501597444089457, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.2299, + "step": 188 + }, + { + "epoch": 0.1509584664536741, + "grad_norm": 0.34375, + "learning_rate": 0.0001, + "loss": 1.2357, + "step": 189 + }, + { + "epoch": 0.15175718849840256, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.2295, + "step": 190 + }, + { + "epoch": 0.152555910543131, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.2409, + "step": 191 + }, + { + "epoch": 0.15335463258785942, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.2221, + "step": 192 + }, + { + "epoch": 0.15415335463258786, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.2363, + "step": 193 + }, + { + "epoch": 0.15495207667731628, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.2348, + "step": 194 + }, + { + "epoch": 0.15575079872204473, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.237, + "step": 195 + }, + { + "epoch": 0.15654952076677317, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.2319, + "step": 196 + }, + { + "epoch": 0.1573482428115016, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.2354, + "step": 197 + }, + { + "epoch": 0.15814696485623003, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.2245, + "step": 198 + }, + { + "epoch": 0.15894568690095848, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.23, + "step": 199 + }, + { + "epoch": 0.1597444089456869, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.2359, + "step": 200 + }, + { + "epoch": 0.16054313099041534, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.2383, + "step": 201 + }, + { + "epoch": 0.16134185303514376, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.2279, + "step": 202 + }, + { + "epoch": 0.1621405750798722, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.2228, + "step": 203 + }, + { + "epoch": 0.16293929712460065, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.2198, + "step": 204 + }, + { + "epoch": 0.16373801916932906, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.2241, + "step": 205 + }, + { + "epoch": 0.1645367412140575, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.2296, + "step": 206 + }, + { + "epoch": 0.16533546325878595, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.223, + "step": 207 + }, + { + "epoch": 0.16613418530351437, + "grad_norm": 0.318359375, + "learning_rate": 0.0001, + "loss": 1.2228, + "step": 208 + }, + { + "epoch": 0.16693290734824281, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.2207, + "step": 209 + }, + { + "epoch": 0.16773162939297126, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.2292, + "step": 210 + }, + { + "epoch": 0.16853035143769968, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.2275, + "step": 211 + }, + { + "epoch": 0.16932907348242812, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.2135, + "step": 212 + }, + { + "epoch": 0.17012779552715654, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.221, + "step": 213 + }, + { + "epoch": 0.17092651757188498, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.219, + "step": 214 + }, + { + "epoch": 0.17172523961661343, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.2172, + "step": 215 + }, + { + "epoch": 0.17252396166134185, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.2183, + "step": 216 + }, + { + "epoch": 0.1733226837060703, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.2235, + "step": 217 + }, + { + "epoch": 0.17412140575079874, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.2115, + "step": 218 + }, + { + "epoch": 0.17492012779552715, + "grad_norm": 0.486328125, + "learning_rate": 0.0001, + "loss": 1.2188, + "step": 219 + }, + { + "epoch": 0.1757188498402556, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.2228, + "step": 220 + }, + { + "epoch": 0.17651757188498401, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.2167, + "step": 221 + }, + { + "epoch": 0.17731629392971246, + "grad_norm": 0.50390625, + "learning_rate": 0.0001, + "loss": 1.2261, + "step": 222 + }, + { + "epoch": 0.1781150159744409, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.2276, + "step": 223 + }, + { + "epoch": 0.17891373801916932, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.2204, + "step": 224 + }, + { + "epoch": 0.17971246006389777, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.2313, + "step": 225 + }, + { + "epoch": 0.1805111821086262, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.2137, + "step": 226 + }, + { + "epoch": 0.18130990415335463, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.2193, + "step": 227 + }, + { + "epoch": 0.18210862619808307, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.2176, + "step": 228 + }, + { + "epoch": 0.1829073482428115, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.2126, + "step": 229 + }, + { + "epoch": 0.18370607028753994, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.223, + "step": 230 + }, + { + "epoch": 0.18450479233226838, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.2147, + "step": 231 + }, + { + "epoch": 0.1853035143769968, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.2261, + "step": 232 + }, + { + "epoch": 0.18610223642172524, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.2189, + "step": 233 + }, + { + "epoch": 0.1869009584664537, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.2262, + "step": 234 + }, + { + "epoch": 0.1876996805111821, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.2173, + "step": 235 + }, + { + "epoch": 0.18849840255591055, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.2194, + "step": 236 + }, + { + "epoch": 0.18929712460063897, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.223, + "step": 237 + }, + { + "epoch": 0.1900958466453674, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.2123, + "step": 238 + }, + { + "epoch": 0.19089456869009586, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.2233, + "step": 239 + }, + { + "epoch": 0.19169329073482427, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.2091, + "step": 240 + }, + { + "epoch": 0.19249201277955272, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.2113, + "step": 241 + }, + { + "epoch": 0.19329073482428116, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.2229, + "step": 242 + }, + { + "epoch": 0.19408945686900958, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.2257, + "step": 243 + }, + { + "epoch": 0.19488817891373802, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.2031, + "step": 244 + }, + { + "epoch": 0.19568690095846644, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.2176, + "step": 245 + }, + { + "epoch": 0.1964856230031949, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.2126, + "step": 246 + }, + { + "epoch": 0.19728434504792333, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.2196, + "step": 247 + }, + { + "epoch": 0.19808306709265175, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.2334, + "step": 248 + }, + { + "epoch": 0.1988817891373802, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.2098, + "step": 249 + }, + { + "epoch": 0.19968051118210864, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.2098, + "step": 250 + }, + { + "epoch": 0.20047923322683706, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.2257, + "step": 251 + }, + { + "epoch": 0.2012779552715655, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.2146, + "step": 252 + }, + { + "epoch": 0.20207667731629392, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.2226, + "step": 253 + }, + { + "epoch": 0.20287539936102236, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.2217, + "step": 254 + }, + { + "epoch": 0.2036741214057508, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.2069, + "step": 255 + }, + { + "epoch": 0.20447284345047922, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.2194, + "step": 256 + }, + { + "epoch": 0.20527156549520767, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.2201, + "step": 257 + }, + { + "epoch": 0.20607028753993611, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.2079, + "step": 258 + }, + { + "epoch": 0.20686900958466453, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1994, + "step": 259 + }, + { + "epoch": 0.20766773162939298, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.207, + "step": 260 + }, + { + "epoch": 0.2084664536741214, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.2117, + "step": 261 + }, + { + "epoch": 0.20926517571884984, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.2141, + "step": 262 + }, + { + "epoch": 0.21006389776357828, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.2106, + "step": 263 + }, + { + "epoch": 0.2108626198083067, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.2171, + "step": 264 + }, + { + "epoch": 0.21166134185303515, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.2153, + "step": 265 + }, + { + "epoch": 0.2124600638977636, + "grad_norm": 0.365234375, + "learning_rate": 0.0001, + "loss": 1.2143, + "step": 266 + }, + { + "epoch": 0.213258785942492, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.2122, + "step": 267 + }, + { + "epoch": 0.21405750798722045, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.2136, + "step": 268 + }, + { + "epoch": 0.21485623003194887, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.206, + "step": 269 + }, + { + "epoch": 0.21565495207667731, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.2107, + "step": 270 + }, + { + "epoch": 0.21645367412140576, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.2042, + "step": 271 + }, + { + "epoch": 0.21725239616613418, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.2096, + "step": 272 + }, + { + "epoch": 0.21805111821086262, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.2156, + "step": 273 + }, + { + "epoch": 0.21884984025559107, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.2143, + "step": 274 + }, + { + "epoch": 0.21964856230031948, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.209, + "step": 275 + }, + { + "epoch": 0.22044728434504793, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.2091, + "step": 276 + }, + { + "epoch": 0.22124600638977635, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.2105, + "step": 277 + }, + { + "epoch": 0.2220447284345048, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.2228, + "step": 278 + }, + { + "epoch": 0.22284345047923323, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.2192, + "step": 279 + }, + { + "epoch": 0.22364217252396165, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.2043, + "step": 280 + }, + { + "epoch": 0.2244408945686901, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.2069, + "step": 281 + }, + { + "epoch": 0.22523961661341854, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.2111, + "step": 282 + }, + { + "epoch": 0.22603833865814696, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.2048, + "step": 283 + }, + { + "epoch": 0.2268370607028754, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1992, + "step": 284 + }, + { + "epoch": 0.22763578274760382, + "grad_norm": 0.36328125, + "learning_rate": 0.0001, + "loss": 1.2094, + "step": 285 + }, + { + "epoch": 0.22843450479233227, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.2071, + "step": 286 + }, + { + "epoch": 0.2292332268370607, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.2037, + "step": 287 + }, + { + "epoch": 0.23003194888178913, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.205, + "step": 288 + }, + { + "epoch": 0.23083067092651757, + "grad_norm": 0.40625, + "learning_rate": 0.0001, + "loss": 1.2043, + "step": 289 + }, + { + "epoch": 0.23162939297124602, + "grad_norm": 0.37890625, + "learning_rate": 0.0001, + "loss": 1.2059, + "step": 290 + }, + { + "epoch": 0.23242811501597443, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.2006, + "step": 291 + }, + { + "epoch": 0.23322683706070288, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.2127, + "step": 292 + }, + { + "epoch": 0.2340255591054313, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.2114, + "step": 293 + }, + { + "epoch": 0.23482428115015974, + "grad_norm": 0.345703125, + "learning_rate": 0.0001, + "loss": 1.21, + "step": 294 + }, + { + "epoch": 0.2356230031948882, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.2086, + "step": 295 + }, + { + "epoch": 0.2364217252396166, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.2, + "step": 296 + }, + { + "epoch": 0.23722044728434505, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1932, + "step": 297 + }, + { + "epoch": 0.2380191693290735, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.2174, + "step": 298 + }, + { + "epoch": 0.2388178913738019, + "grad_norm": 0.34375, + "learning_rate": 0.0001, + "loss": 1.2102, + "step": 299 + }, + { + "epoch": 0.23961661341853036, + "grad_norm": 0.435546875, + "learning_rate": 0.0001, + "loss": 1.2124, + "step": 300 + }, + { + "epoch": 0.24041533546325877, + "grad_norm": 0.8125, + "learning_rate": 0.0001, + "loss": 1.2102, + "step": 301 + }, + { + "epoch": 0.24121405750798722, + "grad_norm": 2.078125, + "learning_rate": 0.0001, + "loss": 1.2284, + "step": 302 + }, + { + "epoch": 0.24201277955271566, + "grad_norm": 0.890625, + "learning_rate": 0.0001, + "loss": 1.2142, + "step": 303 + }, + { + "epoch": 0.24281150159744408, + "grad_norm": 0.8828125, + "learning_rate": 0.0001, + "loss": 1.204, + "step": 304 + }, + { + "epoch": 0.24361022364217252, + "grad_norm": 0.7890625, + "learning_rate": 0.0001, + "loss": 1.2033, + "step": 305 + }, + { + "epoch": 0.24440894568690097, + "grad_norm": 0.53125, + "learning_rate": 0.0001, + "loss": 1.2094, + "step": 306 + }, + { + "epoch": 0.2452076677316294, + "grad_norm": 0.375, + "learning_rate": 0.0001, + "loss": 1.2073, + "step": 307 + }, + { + "epoch": 0.24600638977635783, + "grad_norm": 0.337890625, + "learning_rate": 0.0001, + "loss": 1.2051, + "step": 308 + }, + { + "epoch": 0.24680511182108625, + "grad_norm": 0.453125, + "learning_rate": 0.0001, + "loss": 1.2024, + "step": 309 + }, + { + "epoch": 0.2476038338658147, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.2129, + "step": 310 + }, + { + "epoch": 0.24840255591054314, + "grad_norm": 0.54296875, + "learning_rate": 0.0001, + "loss": 1.2059, + "step": 311 + }, + { + "epoch": 0.24920127795527156, + "grad_norm": 0.50390625, + "learning_rate": 0.0001, + "loss": 1.1947, + "step": 312 + }, + { + "epoch": 0.25, + "grad_norm": 0.431640625, + "learning_rate": 0.0001, + "loss": 1.2161, + "step": 313 + }, + { + "epoch": 0.2507987220447284, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.2084, + "step": 314 + }, + { + "epoch": 0.2515974440894569, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1986, + "step": 315 + }, + { + "epoch": 0.2523961661341853, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1943, + "step": 316 + }, + { + "epoch": 0.2531948881789137, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1971, + "step": 317 + }, + { + "epoch": 0.2539936102236422, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1987, + "step": 318 + }, + { + "epoch": 0.2547923322683706, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.2078, + "step": 319 + }, + { + "epoch": 0.25559105431309903, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.2081, + "step": 320 + }, + { + "epoch": 0.2563897763578275, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.2015, + "step": 321 + }, + { + "epoch": 0.2571884984025559, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.2052, + "step": 322 + }, + { + "epoch": 0.25798722044728434, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.2124, + "step": 323 + }, + { + "epoch": 0.25878594249201275, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1987, + "step": 324 + }, + { + "epoch": 0.2595846645367412, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1945, + "step": 325 + }, + { + "epoch": 0.26038338658146964, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1951, + "step": 326 + }, + { + "epoch": 0.26118210862619806, + "grad_norm": 0.36328125, + "learning_rate": 0.0001, + "loss": 1.2068, + "step": 327 + }, + { + "epoch": 0.26198083067092653, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.2107, + "step": 328 + }, + { + "epoch": 0.26277955271565495, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.2065, + "step": 329 + }, + { + "epoch": 0.26357827476038337, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.2062, + "step": 330 + }, + { + "epoch": 0.26437699680511184, + "grad_norm": 0.345703125, + "learning_rate": 0.0001, + "loss": 1.2117, + "step": 331 + }, + { + "epoch": 0.26517571884984026, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.2095, + "step": 332 + }, + { + "epoch": 0.2659744408945687, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.186, + "step": 333 + }, + { + "epoch": 0.26677316293929715, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1969, + "step": 334 + }, + { + "epoch": 0.26757188498402557, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.2024, + "step": 335 + }, + { + "epoch": 0.268370607028754, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.2061, + "step": 336 + }, + { + "epoch": 0.26916932907348246, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1968, + "step": 337 + }, + { + "epoch": 0.26996805111821087, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.2054, + "step": 338 + }, + { + "epoch": 0.2707667731629393, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1968, + "step": 339 + }, + { + "epoch": 0.2715654952076677, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1957, + "step": 340 + }, + { + "epoch": 0.2723642172523962, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.2144, + "step": 341 + }, + { + "epoch": 0.2731629392971246, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.203, + "step": 342 + }, + { + "epoch": 0.273961661341853, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.2024, + "step": 343 + }, + { + "epoch": 0.2747603833865815, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1979, + "step": 344 + }, + { + "epoch": 0.2755591054313099, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.2057, + "step": 345 + }, + { + "epoch": 0.2763578274760383, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1994, + "step": 346 + }, + { + "epoch": 0.2771565495207668, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.202, + "step": 347 + }, + { + "epoch": 0.2779552715654952, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1937, + "step": 348 + }, + { + "epoch": 0.2787539936102236, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1954, + "step": 349 + }, + { + "epoch": 0.2795527156549521, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.2037, + "step": 350 + }, + { + "epoch": 0.2803514376996805, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1951, + "step": 351 + }, + { + "epoch": 0.28115015974440893, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.198, + "step": 352 + }, + { + "epoch": 0.2819488817891374, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1858, + "step": 353 + }, + { + "epoch": 0.2827476038338658, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.2022, + "step": 354 + }, + { + "epoch": 0.28354632587859424, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1918, + "step": 355 + }, + { + "epoch": 0.28434504792332266, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1955, + "step": 356 + }, + { + "epoch": 0.28514376996805113, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.2058, + "step": 357 + }, + { + "epoch": 0.28594249201277955, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1952, + "step": 358 + }, + { + "epoch": 0.28674121405750796, + "grad_norm": 0.37109375, + "learning_rate": 0.0001, + "loss": 1.195, + "step": 359 + }, + { + "epoch": 0.28753993610223644, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1956, + "step": 360 + }, + { + "epoch": 0.28833865814696485, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.2, + "step": 361 + }, + { + "epoch": 0.28913738019169327, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1918, + "step": 362 + }, + { + "epoch": 0.28993610223642174, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1945, + "step": 363 + }, + { + "epoch": 0.29073482428115016, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1948, + "step": 364 + }, + { + "epoch": 0.2915335463258786, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.2052, + "step": 365 + }, + { + "epoch": 0.29233226837060705, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1942, + "step": 366 + }, + { + "epoch": 0.29313099041533547, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1967, + "step": 367 + }, + { + "epoch": 0.2939297124600639, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.2023, + "step": 368 + }, + { + "epoch": 0.29472843450479236, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1849, + "step": 369 + }, + { + "epoch": 0.2955271565495208, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1929, + "step": 370 + }, + { + "epoch": 0.2963258785942492, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1904, + "step": 371 + }, + { + "epoch": 0.2971246006389776, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.2002, + "step": 372 + }, + { + "epoch": 0.2979233226837061, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.1863, + "step": 373 + }, + { + "epoch": 0.2987220447284345, + "grad_norm": 0.37109375, + "learning_rate": 0.0001, + "loss": 1.1863, + "step": 374 + }, + { + "epoch": 0.2995207667731629, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.2003, + "step": 375 + }, + { + "epoch": 0.3003194888178914, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1993, + "step": 376 + }, + { + "epoch": 0.3011182108626198, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.2052, + "step": 377 + }, + { + "epoch": 0.3019169329073482, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1955, + "step": 378 + }, + { + "epoch": 0.3027156549520767, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1985, + "step": 379 + }, + { + "epoch": 0.3035143769968051, + "grad_norm": 0.380859375, + "learning_rate": 0.0001, + "loss": 1.1987, + "step": 380 + }, + { + "epoch": 0.30431309904153353, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1929, + "step": 381 + }, + { + "epoch": 0.305111821086262, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1954, + "step": 382 + }, + { + "epoch": 0.3059105431309904, + "grad_norm": 0.6015625, + "learning_rate": 0.0001, + "loss": 1.1912, + "step": 383 + }, + { + "epoch": 0.30670926517571884, + "grad_norm": 0.9453125, + "learning_rate": 0.0001, + "loss": 1.1903, + "step": 384 + }, + { + "epoch": 0.3075079872204473, + "grad_norm": 2.03125, + "learning_rate": 0.0001, + "loss": 1.2018, + "step": 385 + }, + { + "epoch": 0.3083067092651757, + "grad_norm": 0.458984375, + "learning_rate": 0.0001, + "loss": 1.1862, + "step": 386 + }, + { + "epoch": 0.30910543130990414, + "grad_norm": 2.5625, + "learning_rate": 0.0001, + "loss": 1.2211, + "step": 387 + }, + { + "epoch": 0.30990415335463256, + "grad_norm": 1.4140625, + "learning_rate": 0.0001, + "loss": 1.2032, + "step": 388 + }, + { + "epoch": 0.31070287539936103, + "grad_norm": 3.421875, + "learning_rate": 0.0001, + "loss": 1.2448, + "step": 389 + }, + { + "epoch": 0.31150159744408945, + "grad_norm": 2.25, + "learning_rate": 0.0001, + "loss": 1.2439, + "step": 390 + }, + { + "epoch": 0.31230031948881787, + "grad_norm": 0.7265625, + "learning_rate": 0.0001, + "loss": 1.2143, + "step": 391 + }, + { + "epoch": 0.31309904153354634, + "grad_norm": 0.80078125, + "learning_rate": 0.0001, + "loss": 1.2182, + "step": 392 + }, + { + "epoch": 0.31389776357827476, + "grad_norm": 1.046875, + "learning_rate": 0.0001, + "loss": 1.23, + "step": 393 + }, + { + "epoch": 0.3146964856230032, + "grad_norm": 1.0078125, + "learning_rate": 0.0001, + "loss": 1.2379, + "step": 394 + }, + { + "epoch": 0.31549520766773165, + "grad_norm": 0.466796875, + "learning_rate": 0.0001, + "loss": 1.2224, + "step": 395 + }, + { + "epoch": 0.31629392971246006, + "grad_norm": 0.5390625, + "learning_rate": 0.0001, + "loss": 1.2137, + "step": 396 + }, + { + "epoch": 0.3170926517571885, + "grad_norm": 0.494140625, + "learning_rate": 0.0001, + "loss": 1.2213, + "step": 397 + }, + { + "epoch": 0.31789137380191695, + "grad_norm": 0.404296875, + "learning_rate": 0.0001, + "loss": 1.2058, + "step": 398 + }, + { + "epoch": 0.31869009584664537, + "grad_norm": 0.37109375, + "learning_rate": 0.0001, + "loss": 1.214, + "step": 399 + }, + { + "epoch": 0.3194888178913738, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.214, + "step": 400 + }, + { + "epoch": 0.32028753993610226, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.2155, + "step": 401 + }, + { + "epoch": 0.3210862619808307, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.2038, + "step": 402 + }, + { + "epoch": 0.3218849840255591, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.2012, + "step": 403 + }, + { + "epoch": 0.3226837060702875, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.2123, + "step": 404 + }, + { + "epoch": 0.323482428115016, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.2026, + "step": 405 + }, + { + "epoch": 0.3242811501597444, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.2087, + "step": 406 + }, + { + "epoch": 0.3250798722044728, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.2042, + "step": 407 + }, + { + "epoch": 0.3258785942492013, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1925, + "step": 408 + }, + { + "epoch": 0.3266773162939297, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.2148, + "step": 409 + }, + { + "epoch": 0.3274760383386581, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1982, + "step": 410 + }, + { + "epoch": 0.3282747603833866, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1975, + "step": 411 + }, + { + "epoch": 0.329073482428115, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.2018, + "step": 412 + }, + { + "epoch": 0.32987220447284343, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1918, + "step": 413 + }, + { + "epoch": 0.3306709265175719, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.2008, + "step": 414 + }, + { + "epoch": 0.3314696485623003, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1941, + "step": 415 + }, + { + "epoch": 0.33226837060702874, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1966, + "step": 416 + }, + { + "epoch": 0.3330670926517572, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1904, + "step": 417 + }, + { + "epoch": 0.33386581469648563, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1999, + "step": 418 + }, + { + "epoch": 0.33466453674121405, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1882, + "step": 419 + }, + { + "epoch": 0.3354632587859425, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1904, + "step": 420 + }, + { + "epoch": 0.33626198083067094, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1906, + "step": 421 + }, + { + "epoch": 0.33706070287539935, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1933, + "step": 422 + }, + { + "epoch": 0.33785942492012777, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1885, + "step": 423 + }, + { + "epoch": 0.33865814696485624, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1909, + "step": 424 + }, + { + "epoch": 0.33945686900958466, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.199, + "step": 425 + }, + { + "epoch": 0.3402555910543131, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1843, + "step": 426 + }, + { + "epoch": 0.34105431309904155, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1819, + "step": 427 + }, + { + "epoch": 0.34185303514376997, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1972, + "step": 428 + }, + { + "epoch": 0.3426517571884984, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1965, + "step": 429 + }, + { + "epoch": 0.34345047923322686, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1922, + "step": 430 + }, + { + "epoch": 0.3442492012779553, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1914, + "step": 431 + }, + { + "epoch": 0.3450479233226837, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.2025, + "step": 432 + }, + { + "epoch": 0.34584664536741216, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1879, + "step": 433 + }, + { + "epoch": 0.3466453674121406, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1857, + "step": 434 + }, + { + "epoch": 0.347444089456869, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1948, + "step": 435 + }, + { + "epoch": 0.34824281150159747, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1974, + "step": 436 + }, + { + "epoch": 0.3490415335463259, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1796, + "step": 437 + }, + { + "epoch": 0.3498402555910543, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1918, + "step": 438 + }, + { + "epoch": 0.3506389776357827, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1873, + "step": 439 + }, + { + "epoch": 0.3514376996805112, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1856, + "step": 440 + }, + { + "epoch": 0.3522364217252396, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1968, + "step": 441 + }, + { + "epoch": 0.35303514376996803, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1901, + "step": 442 + }, + { + "epoch": 0.3538338658146965, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1893, + "step": 443 + }, + { + "epoch": 0.3546325878594249, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1985, + "step": 444 + }, + { + "epoch": 0.35543130990415334, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1875, + "step": 445 + }, + { + "epoch": 0.3562300319488818, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1913, + "step": 446 + }, + { + "epoch": 0.3570287539936102, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1843, + "step": 447 + }, + { + "epoch": 0.35782747603833864, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1925, + "step": 448 + }, + { + "epoch": 0.3586261980830671, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1893, + "step": 449 + }, + { + "epoch": 0.35942492012779553, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1857, + "step": 450 + }, + { + "epoch": 0.36022364217252395, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.2017, + "step": 451 + }, + { + "epoch": 0.3610223642172524, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1937, + "step": 452 + }, + { + "epoch": 0.36182108626198084, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1987, + "step": 453 + }, + { + "epoch": 0.36261980830670926, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1864, + "step": 454 + }, + { + "epoch": 0.3634185303514377, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1875, + "step": 455 + }, + { + "epoch": 0.36421725239616615, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1896, + "step": 456 + }, + { + "epoch": 0.36501597444089456, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1827, + "step": 457 + }, + { + "epoch": 0.365814696485623, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1817, + "step": 458 + }, + { + "epoch": 0.36661341853035145, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1937, + "step": 459 + }, + { + "epoch": 0.36741214057507987, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1822, + "step": 460 + }, + { + "epoch": 0.3682108626198083, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1835, + "step": 461 + }, + { + "epoch": 0.36900958466453676, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1968, + "step": 462 + }, + { + "epoch": 0.3698083067092652, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1941, + "step": 463 + }, + { + "epoch": 0.3706070287539936, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1971, + "step": 464 + }, + { + "epoch": 0.37140575079872207, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1877, + "step": 465 + }, + { + "epoch": 0.3722044728434505, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1764, + "step": 466 + }, + { + "epoch": 0.3730031948881789, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1822, + "step": 467 + }, + { + "epoch": 0.3738019169329074, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1893, + "step": 468 + }, + { + "epoch": 0.3746006389776358, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1822, + "step": 469 + }, + { + "epoch": 0.3753993610223642, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.187, + "step": 470 + }, + { + "epoch": 0.3761980830670926, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1931, + "step": 471 + }, + { + "epoch": 0.3769968051118211, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1888, + "step": 472 + }, + { + "epoch": 0.3777955271565495, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1856, + "step": 473 + }, + { + "epoch": 0.37859424920127793, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1852, + "step": 474 + }, + { + "epoch": 0.3793929712460064, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1816, + "step": 475 + }, + { + "epoch": 0.3801916932907348, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1824, + "step": 476 + }, + { + "epoch": 0.38099041533546324, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1988, + "step": 477 + }, + { + "epoch": 0.3817891373801917, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.183, + "step": 478 + }, + { + "epoch": 0.38258785942492013, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.194, + "step": 479 + }, + { + "epoch": 0.38338658146964855, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1814, + "step": 480 + }, + { + "epoch": 0.384185303514377, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.18, + "step": 481 + }, + { + "epoch": 0.38498402555910544, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1877, + "step": 482 + }, + { + "epoch": 0.38578274760383385, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.188, + "step": 483 + }, + { + "epoch": 0.3865814696485623, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1869, + "step": 484 + }, + { + "epoch": 0.38738019169329074, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1794, + "step": 485 + }, + { + "epoch": 0.38817891373801916, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1913, + "step": 486 + }, + { + "epoch": 0.3889776357827476, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1727, + "step": 487 + }, + { + "epoch": 0.38977635782747605, + "grad_norm": 0.361328125, + "learning_rate": 0.0001, + "loss": 1.188, + "step": 488 + }, + { + "epoch": 0.39057507987220447, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1815, + "step": 489 + }, + { + "epoch": 0.3913738019169329, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1762, + "step": 490 + }, + { + "epoch": 0.39217252396166136, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1866, + "step": 491 + }, + { + "epoch": 0.3929712460063898, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.187, + "step": 492 + }, + { + "epoch": 0.3937699680511182, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1944, + "step": 493 + }, + { + "epoch": 0.39456869009584666, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1885, + "step": 494 + }, + { + "epoch": 0.3953674121405751, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1867, + "step": 495 + }, + { + "epoch": 0.3961661341853035, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1764, + "step": 496 + }, + { + "epoch": 0.39696485623003197, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.176, + "step": 497 + }, + { + "epoch": 0.3977635782747604, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.181, + "step": 498 + }, + { + "epoch": 0.3985623003194888, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1896, + "step": 499 + }, + { + "epoch": 0.3993610223642173, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1897, + "step": 500 + }, + { + "epoch": 0.4001597444089457, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1804, + "step": 501 + }, + { + "epoch": 0.4009584664536741, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1882, + "step": 502 + }, + { + "epoch": 0.40175718849840253, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1901, + "step": 503 + }, + { + "epoch": 0.402555910543131, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1888, + "step": 504 + }, + { + "epoch": 0.4033546325878594, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1865, + "step": 505 + }, + { + "epoch": 0.40415335463258784, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1838, + "step": 506 + }, + { + "epoch": 0.4049520766773163, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1735, + "step": 507 + }, + { + "epoch": 0.4057507987220447, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1654, + "step": 508 + }, + { + "epoch": 0.40654952076677314, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1912, + "step": 509 + }, + { + "epoch": 0.4073482428115016, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.182, + "step": 510 + }, + { + "epoch": 0.40814696485623003, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1897, + "step": 511 + }, + { + "epoch": 0.40894568690095845, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1882, + "step": 512 + }, + { + "epoch": 0.4097444089456869, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1785, + "step": 513 + }, + { + "epoch": 0.41054313099041534, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1775, + "step": 514 + }, + { + "epoch": 0.41134185303514376, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.182, + "step": 515 + }, + { + "epoch": 0.41214057507987223, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1808, + "step": 516 + }, + { + "epoch": 0.41293929712460065, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1856, + "step": 517 + }, + { + "epoch": 0.41373801916932906, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1803, + "step": 518 + }, + { + "epoch": 0.4145367412140575, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1762, + "step": 519 + }, + { + "epoch": 0.41533546325878595, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1836, + "step": 520 + }, + { + "epoch": 0.41613418530351437, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1831, + "step": 521 + }, + { + "epoch": 0.4169329073482428, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.174, + "step": 522 + }, + { + "epoch": 0.41773162939297126, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.179, + "step": 523 + }, + { + "epoch": 0.4185303514376997, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1816, + "step": 524 + }, + { + "epoch": 0.4193290734824281, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1831, + "step": 525 + }, + { + "epoch": 0.42012779552715657, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1848, + "step": 526 + }, + { + "epoch": 0.420926517571885, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1804, + "step": 527 + }, + { + "epoch": 0.4217252396166134, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1871, + "step": 528 + }, + { + "epoch": 0.4225239616613419, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1752, + "step": 529 + }, + { + "epoch": 0.4233226837060703, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.181, + "step": 530 + }, + { + "epoch": 0.4241214057507987, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1846, + "step": 531 + }, + { + "epoch": 0.4249201277955272, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1796, + "step": 532 + }, + { + "epoch": 0.4257188498402556, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1701, + "step": 533 + }, + { + "epoch": 0.426517571884984, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1774, + "step": 534 + }, + { + "epoch": 0.4273162939297125, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1731, + "step": 535 + }, + { + "epoch": 0.4281150159744409, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1843, + "step": 536 + }, + { + "epoch": 0.4289137380191693, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1902, + "step": 537 + }, + { + "epoch": 0.42971246006389774, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1865, + "step": 538 + }, + { + "epoch": 0.4305111821086262, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1827, + "step": 539 + }, + { + "epoch": 0.43130990415335463, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.177, + "step": 540 + }, + { + "epoch": 0.43210862619808305, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1757, + "step": 541 + }, + { + "epoch": 0.4329073482428115, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1864, + "step": 542 + }, + { + "epoch": 0.43370607028753994, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1814, + "step": 543 + }, + { + "epoch": 0.43450479233226835, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1977, + "step": 544 + }, + { + "epoch": 0.4353035143769968, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1782, + "step": 545 + }, + { + "epoch": 0.43610223642172524, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1805, + "step": 546 + }, + { + "epoch": 0.43690095846645366, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1757, + "step": 547 + }, + { + "epoch": 0.43769968051118213, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1818, + "step": 548 + }, + { + "epoch": 0.43849840255591055, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1848, + "step": 549 + }, + { + "epoch": 0.43929712460063897, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1747, + "step": 550 + }, + { + "epoch": 0.44009584664536744, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.1817, + "step": 551 + }, + { + "epoch": 0.44089456869009586, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1882, + "step": 552 + }, + { + "epoch": 0.4416932907348243, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1784, + "step": 553 + }, + { + "epoch": 0.4424920127795527, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1803, + "step": 554 + }, + { + "epoch": 0.44329073482428116, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1793, + "step": 555 + }, + { + "epoch": 0.4440894568690096, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1636, + "step": 556 + }, + { + "epoch": 0.444888178913738, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1758, + "step": 557 + }, + { + "epoch": 0.44568690095846647, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1743, + "step": 558 + }, + { + "epoch": 0.4464856230031949, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1791, + "step": 559 + }, + { + "epoch": 0.4472843450479233, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1781, + "step": 560 + }, + { + "epoch": 0.4480830670926518, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1754, + "step": 561 + }, + { + "epoch": 0.4488817891373802, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1759, + "step": 562 + }, + { + "epoch": 0.4496805111821086, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1831, + "step": 563 + }, + { + "epoch": 0.4504792332268371, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1874, + "step": 564 + }, + { + "epoch": 0.4512779552715655, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1777, + "step": 565 + }, + { + "epoch": 0.4520766773162939, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1797, + "step": 566 + }, + { + "epoch": 0.4528753993610224, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1806, + "step": 567 + }, + { + "epoch": 0.4536741214057508, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.177, + "step": 568 + }, + { + "epoch": 0.4544728434504792, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1709, + "step": 569 + }, + { + "epoch": 0.45527156549520764, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1815, + "step": 570 + }, + { + "epoch": 0.4560702875399361, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1861, + "step": 571 + }, + { + "epoch": 0.45686900958466453, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1652, + "step": 572 + }, + { + "epoch": 0.45766773162939295, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.176, + "step": 573 + }, + { + "epoch": 0.4584664536741214, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1846, + "step": 574 + }, + { + "epoch": 0.45926517571884984, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1663, + "step": 575 + }, + { + "epoch": 0.46006389776357826, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1808, + "step": 576 + }, + { + "epoch": 0.46086261980830673, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1747, + "step": 577 + }, + { + "epoch": 0.46166134185303515, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1735, + "step": 578 + }, + { + "epoch": 0.46246006389776356, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1746, + "step": 579 + }, + { + "epoch": 0.46325878594249204, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1772, + "step": 580 + }, + { + "epoch": 0.46405750798722045, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1754, + "step": 581 + }, + { + "epoch": 0.46485623003194887, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1819, + "step": 582 + }, + { + "epoch": 0.46565495207667734, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1803, + "step": 583 + }, + { + "epoch": 0.46645367412140576, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1739, + "step": 584 + }, + { + "epoch": 0.4672523961661342, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1817, + "step": 585 + }, + { + "epoch": 0.4680511182108626, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1803, + "step": 586 + }, + { + "epoch": 0.46884984025559107, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1741, + "step": 587 + }, + { + "epoch": 0.4696485623003195, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.177, + "step": 588 + }, + { + "epoch": 0.4704472843450479, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1753, + "step": 589 + }, + { + "epoch": 0.4712460063897764, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1792, + "step": 590 + }, + { + "epoch": 0.4720447284345048, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1653, + "step": 591 + }, + { + "epoch": 0.4728434504792332, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1752, + "step": 592 + }, + { + "epoch": 0.4736421725239617, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1808, + "step": 593 + }, + { + "epoch": 0.4744408945686901, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1726, + "step": 594 + }, + { + "epoch": 0.4752396166134185, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1734, + "step": 595 + }, + { + "epoch": 0.476038338658147, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.182, + "step": 596 + }, + { + "epoch": 0.4768370607028754, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1776, + "step": 597 + }, + { + "epoch": 0.4776357827476038, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1729, + "step": 598 + }, + { + "epoch": 0.4784345047923323, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1838, + "step": 599 + }, + { + "epoch": 0.4792332268370607, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1731, + "step": 600 + }, + { + "epoch": 0.48003194888178913, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1816, + "step": 601 + }, + { + "epoch": 0.48083067092651754, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.182, + "step": 602 + }, + { + "epoch": 0.481629392971246, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.1708, + "step": 603 + }, + { + "epoch": 0.48242811501597443, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1721, + "step": 604 + }, + { + "epoch": 0.48322683706070285, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1757, + "step": 605 + }, + { + "epoch": 0.4840255591054313, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1648, + "step": 606 + }, + { + "epoch": 0.48482428115015974, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1695, + "step": 607 + }, + { + "epoch": 0.48562300319488816, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1784, + "step": 608 + }, + { + "epoch": 0.48642172523961663, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1928, + "step": 609 + }, + { + "epoch": 0.48722044728434505, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1784, + "step": 610 + }, + { + "epoch": 0.48801916932907347, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.176, + "step": 611 + }, + { + "epoch": 0.48881789137380194, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1802, + "step": 612 + }, + { + "epoch": 0.48961661341853036, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1784, + "step": 613 + }, + { + "epoch": 0.4904153354632588, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1742, + "step": 614 + }, + { + "epoch": 0.49121405750798725, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1798, + "step": 615 + }, + { + "epoch": 0.49201277955271566, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1755, + "step": 616 + }, + { + "epoch": 0.4928115015974441, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.182, + "step": 617 + }, + { + "epoch": 0.4936102236421725, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1675, + "step": 618 + }, + { + "epoch": 0.49440894568690097, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1831, + "step": 619 + }, + { + "epoch": 0.4952076677316294, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1682, + "step": 620 + }, + { + "epoch": 0.4960063897763578, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1707, + "step": 621 + }, + { + "epoch": 0.4968051118210863, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1711, + "step": 622 + }, + { + "epoch": 0.4976038338658147, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1755, + "step": 623 + }, + { + "epoch": 0.4984025559105431, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1698, + "step": 624 + }, + { + "epoch": 0.4992012779552716, + "grad_norm": 0.419921875, + "learning_rate": 0.0001, + "loss": 1.1776, + "step": 625 + }, + { + "epoch": 0.5, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1853, + "step": 626 + }, + { + "epoch": 0.5007987220447284, + "grad_norm": 0.37890625, + "learning_rate": 0.0001, + "loss": 1.1829, + "step": 627 + }, + { + "epoch": 0.5015974440894568, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1737, + "step": 628 + }, + { + "epoch": 0.5023961661341853, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.174, + "step": 629 + }, + { + "epoch": 0.5031948881789138, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1732, + "step": 630 + }, + { + "epoch": 0.5039936102236422, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1786, + "step": 631 + }, + { + "epoch": 0.5047923322683706, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1849, + "step": 632 + }, + { + "epoch": 0.505591054313099, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1729, + "step": 633 + }, + { + "epoch": 0.5063897763578274, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.177, + "step": 634 + }, + { + "epoch": 0.5071884984025559, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.1832, + "step": 635 + }, + { + "epoch": 0.5079872204472844, + "grad_norm": 0.462890625, + "learning_rate": 0.0001, + "loss": 1.1745, + "step": 636 + }, + { + "epoch": 0.5087859424920128, + "grad_norm": 0.38671875, + "learning_rate": 0.0001, + "loss": 1.1792, + "step": 637 + }, + { + "epoch": 0.5095846645367412, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1694, + "step": 638 + }, + { + "epoch": 0.5103833865814696, + "grad_norm": 0.36328125, + "learning_rate": 0.0001, + "loss": 1.1844, + "step": 639 + }, + { + "epoch": 0.5111821086261981, + "grad_norm": 0.380859375, + "learning_rate": 0.0001, + "loss": 1.1775, + "step": 640 + }, + { + "epoch": 0.5119808306709265, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.19, + "step": 641 + }, + { + "epoch": 0.512779552715655, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1783, + "step": 642 + }, + { + "epoch": 0.5135782747603834, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1813, + "step": 643 + }, + { + "epoch": 0.5143769968051118, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.171, + "step": 644 + }, + { + "epoch": 0.5151757188498403, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1733, + "step": 645 + }, + { + "epoch": 0.5159744408945687, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1842, + "step": 646 + }, + { + "epoch": 0.5167731629392971, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1835, + "step": 647 + }, + { + "epoch": 0.5175718849840255, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1832, + "step": 648 + }, + { + "epoch": 0.518370607028754, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1799, + "step": 649 + }, + { + "epoch": 0.5191693290734825, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1837, + "step": 650 + }, + { + "epoch": 0.5199680511182109, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.178, + "step": 651 + }, + { + "epoch": 0.5207667731629393, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1645, + "step": 652 + }, + { + "epoch": 0.5215654952076677, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1654, + "step": 653 + }, + { + "epoch": 0.5223642172523961, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1691, + "step": 654 + }, + { + "epoch": 0.5231629392971247, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1729, + "step": 655 + }, + { + "epoch": 0.5239616613418531, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1634, + "step": 656 + }, + { + "epoch": 0.5247603833865815, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1669, + "step": 657 + }, + { + "epoch": 0.5255591054313099, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1798, + "step": 658 + }, + { + "epoch": 0.5263578274760383, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1716, + "step": 659 + }, + { + "epoch": 0.5271565495207667, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1837, + "step": 660 + }, + { + "epoch": 0.5279552715654952, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1716, + "step": 661 + }, + { + "epoch": 0.5287539936102237, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1863, + "step": 662 + }, + { + "epoch": 0.5295527156549521, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1673, + "step": 663 + }, + { + "epoch": 0.5303514376996805, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1672, + "step": 664 + }, + { + "epoch": 0.5311501597444089, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1732, + "step": 665 + }, + { + "epoch": 0.5319488817891374, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1593, + "step": 666 + }, + { + "epoch": 0.5327476038338658, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.17, + "step": 667 + }, + { + "epoch": 0.5335463258785943, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1764, + "step": 668 + }, + { + "epoch": 0.5343450479233227, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.166, + "step": 669 + }, + { + "epoch": 0.5351437699680511, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1752, + "step": 670 + }, + { + "epoch": 0.5359424920127795, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1648, + "step": 671 + }, + { + "epoch": 0.536741214057508, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1684, + "step": 672 + }, + { + "epoch": 0.5375399361022364, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1679, + "step": 673 + }, + { + "epoch": 0.5383386581469649, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1683, + "step": 674 + }, + { + "epoch": 0.5391373801916933, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1721, + "step": 675 + }, + { + "epoch": 0.5399361022364217, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1736, + "step": 676 + }, + { + "epoch": 0.5407348242811502, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1875, + "step": 677 + }, + { + "epoch": 0.5415335463258786, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.179, + "step": 678 + }, + { + "epoch": 0.542332268370607, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.169, + "step": 679 + }, + { + "epoch": 0.5431309904153354, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.168, + "step": 680 + }, + { + "epoch": 0.5439297124600639, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.178, + "step": 681 + }, + { + "epoch": 0.5447284345047924, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1781, + "step": 682 + }, + { + "epoch": 0.5455271565495208, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.172, + "step": 683 + }, + { + "epoch": 0.5463258785942492, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1761, + "step": 684 + }, + { + "epoch": 0.5471246006389776, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1707, + "step": 685 + }, + { + "epoch": 0.547923322683706, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1705, + "step": 686 + }, + { + "epoch": 0.5487220447284346, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1725, + "step": 687 + }, + { + "epoch": 0.549520766773163, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1686, + "step": 688 + }, + { + "epoch": 0.5503194888178914, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1673, + "step": 689 + }, + { + "epoch": 0.5511182108626198, + "grad_norm": 0.41015625, + "learning_rate": 0.0001, + "loss": 1.176, + "step": 690 + }, + { + "epoch": 0.5519169329073482, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1747, + "step": 691 + }, + { + "epoch": 0.5527156549520766, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1601, + "step": 692 + }, + { + "epoch": 0.5535143769968051, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1708, + "step": 693 + }, + { + "epoch": 0.5543130990415336, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1821, + "step": 694 + }, + { + "epoch": 0.555111821086262, + "grad_norm": 0.39453125, + "learning_rate": 0.0001, + "loss": 1.1665, + "step": 695 + }, + { + "epoch": 0.5559105431309904, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1715, + "step": 696 + }, + { + "epoch": 0.5567092651757188, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1811, + "step": 697 + }, + { + "epoch": 0.5575079872204473, + "grad_norm": 0.37109375, + "learning_rate": 0.0001, + "loss": 1.1732, + "step": 698 + }, + { + "epoch": 0.5583067092651757, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1795, + "step": 699 + }, + { + "epoch": 0.5591054313099042, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1631, + "step": 700 + }, + { + "epoch": 0.5599041533546326, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1658, + "step": 701 + }, + { + "epoch": 0.560702875399361, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1668, + "step": 702 + }, + { + "epoch": 0.5615015974440895, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1657, + "step": 703 + }, + { + "epoch": 0.5623003194888179, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1756, + "step": 704 + }, + { + "epoch": 0.5630990415335463, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1735, + "step": 705 + }, + { + "epoch": 0.5638977635782748, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.178, + "step": 706 + }, + { + "epoch": 0.5646964856230032, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1598, + "step": 707 + }, + { + "epoch": 0.5654952076677316, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1709, + "step": 708 + }, + { + "epoch": 0.5662939297124601, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1645, + "step": 709 + }, + { + "epoch": 0.5670926517571885, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1748, + "step": 710 + }, + { + "epoch": 0.5678913738019169, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1738, + "step": 711 + }, + { + "epoch": 0.5686900958466453, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1582, + "step": 712 + }, + { + "epoch": 0.5694888178913738, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1703, + "step": 713 + }, + { + "epoch": 0.5702875399361023, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1598, + "step": 714 + }, + { + "epoch": 0.5710862619808307, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1721, + "step": 715 + }, + { + "epoch": 0.5718849840255591, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1603, + "step": 716 + }, + { + "epoch": 0.5726837060702875, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1625, + "step": 717 + }, + { + "epoch": 0.5734824281150159, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1858, + "step": 718 + }, + { + "epoch": 0.5742811501597445, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1621, + "step": 719 + }, + { + "epoch": 0.5750798722044729, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.164, + "step": 720 + }, + { + "epoch": 0.5758785942492013, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1762, + "step": 721 + }, + { + "epoch": 0.5766773162939297, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1715, + "step": 722 + }, + { + "epoch": 0.5774760383386581, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1671, + "step": 723 + }, + { + "epoch": 0.5782747603833865, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1794, + "step": 724 + }, + { + "epoch": 0.579073482428115, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1659, + "step": 725 + }, + { + "epoch": 0.5798722044728435, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1789, + "step": 726 + }, + { + "epoch": 0.5806709265175719, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1769, + "step": 727 + }, + { + "epoch": 0.5814696485623003, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1609, + "step": 728 + }, + { + "epoch": 0.5822683706070287, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1609, + "step": 729 + }, + { + "epoch": 0.5830670926517572, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.159, + "step": 730 + }, + { + "epoch": 0.5838658146964856, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1784, + "step": 731 + }, + { + "epoch": 0.5846645367412141, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1673, + "step": 732 + }, + { + "epoch": 0.5854632587859425, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1702, + "step": 733 + }, + { + "epoch": 0.5862619808306709, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1704, + "step": 734 + }, + { + "epoch": 0.5870607028753994, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1619, + "step": 735 + }, + { + "epoch": 0.5878594249201278, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.173, + "step": 736 + }, + { + "epoch": 0.5886581469648562, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1696, + "step": 737 + }, + { + "epoch": 0.5894568690095847, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1714, + "step": 738 + }, + { + "epoch": 0.5902555910543131, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1635, + "step": 739 + }, + { + "epoch": 0.5910543130990416, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1597, + "step": 740 + }, + { + "epoch": 0.59185303514377, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1735, + "step": 741 + }, + { + "epoch": 0.5926517571884984, + "grad_norm": 0.3828125, + "learning_rate": 0.0001, + "loss": 1.1654, + "step": 742 + }, + { + "epoch": 0.5934504792332268, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1574, + "step": 743 + }, + { + "epoch": 0.5942492012779552, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.179, + "step": 744 + }, + { + "epoch": 0.5950479233226837, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1586, + "step": 745 + }, + { + "epoch": 0.5958466453674122, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1732, + "step": 746 + }, + { + "epoch": 0.5966453674121406, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1686, + "step": 747 + }, + { + "epoch": 0.597444089456869, + "grad_norm": 0.54296875, + "learning_rate": 0.0001, + "loss": 1.1642, + "step": 748 + }, + { + "epoch": 0.5982428115015974, + "grad_norm": 0.671875, + "learning_rate": 0.0001, + "loss": 1.1629, + "step": 749 + }, + { + "epoch": 0.5990415335463258, + "grad_norm": 0.80078125, + "learning_rate": 0.0001, + "loss": 1.1672, + "step": 750 + }, + { + "epoch": 0.5998402555910544, + "grad_norm": 1.0546875, + "learning_rate": 0.0001, + "loss": 1.173, + "step": 751 + }, + { + "epoch": 0.6006389776357828, + "grad_norm": 1.6484375, + "learning_rate": 0.0001, + "loss": 1.1713, + "step": 752 + }, + { + "epoch": 0.6014376996805112, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1711, + "step": 753 + }, + { + "epoch": 0.6022364217252396, + "grad_norm": 1.59375, + "learning_rate": 0.0001, + "loss": 1.1793, + "step": 754 + }, + { + "epoch": 0.603035143769968, + "grad_norm": 0.86328125, + "learning_rate": 0.0001, + "loss": 1.1794, + "step": 755 + }, + { + "epoch": 0.6038338658146964, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1639, + "step": 756 + }, + { + "epoch": 0.604632587859425, + "grad_norm": 1.0390625, + "learning_rate": 0.0001, + "loss": 1.175, + "step": 757 + }, + { + "epoch": 0.6054313099041534, + "grad_norm": 1.2109375, + "learning_rate": 0.0001, + "loss": 1.1752, + "step": 758 + }, + { + "epoch": 0.6062300319488818, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.1733, + "step": 759 + }, + { + "epoch": 0.6070287539936102, + "grad_norm": 1.0625, + "learning_rate": 0.0001, + "loss": 1.1853, + "step": 760 + }, + { + "epoch": 0.6078274760383386, + "grad_norm": 1.7421875, + "learning_rate": 0.0001, + "loss": 1.1644, + "step": 761 + }, + { + "epoch": 0.6086261980830671, + "grad_norm": 0.72265625, + "learning_rate": 0.0001, + "loss": 1.1848, + "step": 762 + }, + { + "epoch": 0.6094249201277955, + "grad_norm": 2.796875, + "learning_rate": 0.0001, + "loss": 1.1993, + "step": 763 + }, + { + "epoch": 0.610223642172524, + "grad_norm": 1.96875, + "learning_rate": 0.0001, + "loss": 1.2001, + "step": 764 + }, + { + "epoch": 0.6110223642172524, + "grad_norm": 1.71875, + "learning_rate": 0.0001, + "loss": 1.1999, + "step": 765 + }, + { + "epoch": 0.6118210862619808, + "grad_norm": 1.4140625, + "learning_rate": 0.0001, + "loss": 1.1795, + "step": 766 + }, + { + "epoch": 0.6126198083067093, + "grad_norm": 0.98828125, + "learning_rate": 0.0001, + "loss": 1.1849, + "step": 767 + }, + { + "epoch": 0.6134185303514377, + "grad_norm": 0.6875, + "learning_rate": 0.0001, + "loss": 1.1698, + "step": 768 + }, + { + "epoch": 0.6142172523961661, + "grad_norm": 0.515625, + "learning_rate": 0.0001, + "loss": 1.1816, + "step": 769 + }, + { + "epoch": 0.6150159744408946, + "grad_norm": 0.5234375, + "learning_rate": 0.0001, + "loss": 1.1838, + "step": 770 + }, + { + "epoch": 0.615814696485623, + "grad_norm": 0.498046875, + "learning_rate": 0.0001, + "loss": 1.1835, + "step": 771 + }, + { + "epoch": 0.6166134185303515, + "grad_norm": 0.470703125, + "learning_rate": 0.0001, + "loss": 1.1733, + "step": 772 + }, + { + "epoch": 0.6174121405750799, + "grad_norm": 0.421875, + "learning_rate": 0.0001, + "loss": 1.172, + "step": 773 + }, + { + "epoch": 0.6182108626198083, + "grad_norm": 0.462890625, + "learning_rate": 0.0001, + "loss": 1.1755, + "step": 774 + }, + { + "epoch": 0.6190095846645367, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.1855, + "step": 775 + }, + { + "epoch": 0.6198083067092651, + "grad_norm": 0.373046875, + "learning_rate": 0.0001, + "loss": 1.1738, + "step": 776 + }, + { + "epoch": 0.6206070287539937, + "grad_norm": 0.427734375, + "learning_rate": 0.0001, + "loss": 1.1766, + "step": 777 + }, + { + "epoch": 0.6214057507987221, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1802, + "step": 778 + }, + { + "epoch": 0.6222044728434505, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1723, + "step": 779 + }, + { + "epoch": 0.6230031948881789, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1778, + "step": 780 + }, + { + "epoch": 0.6238019169329073, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1594, + "step": 781 + }, + { + "epoch": 0.6246006389776357, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1808, + "step": 782 + }, + { + "epoch": 0.6253993610223643, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1849, + "step": 783 + }, + { + "epoch": 0.6261980830670927, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1725, + "step": 784 + }, + { + "epoch": 0.6269968051118211, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1707, + "step": 785 + }, + { + "epoch": 0.6277955271565495, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1786, + "step": 786 + }, + { + "epoch": 0.6285942492012779, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1693, + "step": 787 + }, + { + "epoch": 0.6293929712460063, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1758, + "step": 788 + }, + { + "epoch": 0.6301916932907349, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1698, + "step": 789 + }, + { + "epoch": 0.6309904153354633, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1585, + "step": 790 + }, + { + "epoch": 0.6317891373801917, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.169, + "step": 791 + }, + { + "epoch": 0.6325878594249201, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.17, + "step": 792 + }, + { + "epoch": 0.6333865814696485, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1697, + "step": 793 + }, + { + "epoch": 0.634185303514377, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1719, + "step": 794 + }, + { + "epoch": 0.6349840255591054, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1601, + "step": 795 + }, + { + "epoch": 0.6357827476038339, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1713, + "step": 796 + }, + { + "epoch": 0.6365814696485623, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1605, + "step": 797 + }, + { + "epoch": 0.6373801916932907, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1635, + "step": 798 + }, + { + "epoch": 0.6381789137380192, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1785, + "step": 799 + }, + { + "epoch": 0.6389776357827476, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1729, + "step": 800 + }, + { + "epoch": 0.639776357827476, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.169, + "step": 801 + }, + { + "epoch": 0.6405750798722045, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1662, + "step": 802 + }, + { + "epoch": 0.6413738019169329, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1637, + "step": 803 + }, + { + "epoch": 0.6421725239616614, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1786, + "step": 804 + }, + { + "epoch": 0.6429712460063898, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1598, + "step": 805 + }, + { + "epoch": 0.6437699680511182, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1571, + "step": 806 + }, + { + "epoch": 0.6445686900958466, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1722, + "step": 807 + }, + { + "epoch": 0.645367412140575, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1593, + "step": 808 + }, + { + "epoch": 0.6461661341853036, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1621, + "step": 809 + }, + { + "epoch": 0.646964856230032, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1678, + "step": 810 + }, + { + "epoch": 0.6477635782747604, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1675, + "step": 811 + }, + { + "epoch": 0.6485623003194888, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1745, + "step": 812 + }, + { + "epoch": 0.6493610223642172, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1649, + "step": 813 + }, + { + "epoch": 0.6501597444089456, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1671, + "step": 814 + }, + { + "epoch": 0.6509584664536742, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1639, + "step": 815 + }, + { + "epoch": 0.6517571884984026, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1687, + "step": 816 + }, + { + "epoch": 0.652555910543131, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1722, + "step": 817 + }, + { + "epoch": 0.6533546325878594, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1582, + "step": 818 + }, + { + "epoch": 0.6541533546325878, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1647, + "step": 819 + }, + { + "epoch": 0.6549520766773163, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1628, + "step": 820 + }, + { + "epoch": 0.6557507987220448, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1599, + "step": 821 + }, + { + "epoch": 0.6565495207667732, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1604, + "step": 822 + }, + { + "epoch": 0.6573482428115016, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1676, + "step": 823 + }, + { + "epoch": 0.65814696485623, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1662, + "step": 824 + }, + { + "epoch": 0.6589456869009584, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.164, + "step": 825 + }, + { + "epoch": 0.6597444089456869, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1726, + "step": 826 + }, + { + "epoch": 0.6605431309904153, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1655, + "step": 827 + }, + { + "epoch": 0.6613418530351438, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1663, + "step": 828 + }, + { + "epoch": 0.6621405750798722, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1595, + "step": 829 + }, + { + "epoch": 0.6629392971246006, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1705, + "step": 830 + }, + { + "epoch": 0.6637380191693291, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.164, + "step": 831 + }, + { + "epoch": 0.6645367412140575, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1646, + "step": 832 + }, + { + "epoch": 0.6653354632587859, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1741, + "step": 833 + }, + { + "epoch": 0.6661341853035144, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1681, + "step": 834 + }, + { + "epoch": 0.6669329073482428, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1708, + "step": 835 + }, + { + "epoch": 0.6677316293929713, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1684, + "step": 836 + }, + { + "epoch": 0.6685303514376997, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1659, + "step": 837 + }, + { + "epoch": 0.6693290734824281, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.172, + "step": 838 + }, + { + "epoch": 0.6701277955271565, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.154, + "step": 839 + }, + { + "epoch": 0.670926517571885, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1594, + "step": 840 + }, + { + "epoch": 0.6717252396166135, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1627, + "step": 841 + }, + { + "epoch": 0.6725239616613419, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1611, + "step": 842 + }, + { + "epoch": 0.6733226837060703, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.16, + "step": 843 + }, + { + "epoch": 0.6741214057507987, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1729, + "step": 844 + }, + { + "epoch": 0.6749201277955271, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.151, + "step": 845 + }, + { + "epoch": 0.6757188498402555, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1705, + "step": 846 + }, + { + "epoch": 0.6765175718849841, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1583, + "step": 847 + }, + { + "epoch": 0.6773162939297125, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1653, + "step": 848 + }, + { + "epoch": 0.6781150159744409, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.158, + "step": 849 + }, + { + "epoch": 0.6789137380191693, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1561, + "step": 850 + }, + { + "epoch": 0.6797124600638977, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1685, + "step": 851 + }, + { + "epoch": 0.6805111821086262, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1653, + "step": 852 + }, + { + "epoch": 0.6813099041533547, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1583, + "step": 853 + }, + { + "epoch": 0.6821086261980831, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1604, + "step": 854 + }, + { + "epoch": 0.6829073482428115, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1683, + "step": 855 + }, + { + "epoch": 0.6837060702875399, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1613, + "step": 856 + }, + { + "epoch": 0.6845047923322684, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.163, + "step": 857 + }, + { + "epoch": 0.6853035143769968, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1566, + "step": 858 + }, + { + "epoch": 0.6861022364217252, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1617, + "step": 859 + }, + { + "epoch": 0.6869009584664537, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1624, + "step": 860 + }, + { + "epoch": 0.6876996805111821, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1635, + "step": 861 + }, + { + "epoch": 0.6884984025559105, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1656, + "step": 862 + }, + { + "epoch": 0.689297124600639, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1551, + "step": 863 + }, + { + "epoch": 0.6900958466453674, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1571, + "step": 864 + }, + { + "epoch": 0.6908945686900958, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1713, + "step": 865 + }, + { + "epoch": 0.6916932907348243, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1643, + "step": 866 + }, + { + "epoch": 0.6924920127795527, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1656, + "step": 867 + }, + { + "epoch": 0.6932907348242812, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1626, + "step": 868 + }, + { + "epoch": 0.6940894568690096, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1587, + "step": 869 + }, + { + "epoch": 0.694888178913738, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1505, + "step": 870 + }, + { + "epoch": 0.6956869009584664, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1583, + "step": 871 + }, + { + "epoch": 0.6964856230031949, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1699, + "step": 872 + }, + { + "epoch": 0.6972843450479234, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1565, + "step": 873 + }, + { + "epoch": 0.6980830670926518, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1672, + "step": 874 + }, + { + "epoch": 0.6988817891373802, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.162, + "step": 875 + }, + { + "epoch": 0.6996805111821086, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1569, + "step": 876 + }, + { + "epoch": 0.700479233226837, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1689, + "step": 877 + }, + { + "epoch": 0.7012779552715654, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1603, + "step": 878 + }, + { + "epoch": 0.702076677316294, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.158, + "step": 879 + }, + { + "epoch": 0.7028753993610224, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1609, + "step": 880 + }, + { + "epoch": 0.7036741214057508, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.161, + "step": 881 + }, + { + "epoch": 0.7044728434504792, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1712, + "step": 882 + }, + { + "epoch": 0.7052715654952076, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.168, + "step": 883 + }, + { + "epoch": 0.7060702875399361, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1738, + "step": 884 + }, + { + "epoch": 0.7068690095846646, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1574, + "step": 885 + }, + { + "epoch": 0.707667731629393, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1639, + "step": 886 + }, + { + "epoch": 0.7084664536741214, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1598, + "step": 887 + }, + { + "epoch": 0.7092651757188498, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1624, + "step": 888 + }, + { + "epoch": 0.7100638977635783, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1567, + "step": 889 + }, + { + "epoch": 0.7108626198083067, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1606, + "step": 890 + }, + { + "epoch": 0.7116613418530351, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1596, + "step": 891 + }, + { + "epoch": 0.7124600638977636, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1615, + "step": 892 + }, + { + "epoch": 0.713258785942492, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1638, + "step": 893 + }, + { + "epoch": 0.7140575079872205, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1646, + "step": 894 + }, + { + "epoch": 0.7148562300319489, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1645, + "step": 895 + }, + { + "epoch": 0.7156549520766773, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1603, + "step": 896 + }, + { + "epoch": 0.7164536741214057, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1678, + "step": 897 + }, + { + "epoch": 0.7172523961661342, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1547, + "step": 898 + }, + { + "epoch": 0.7180511182108626, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1681, + "step": 899 + }, + { + "epoch": 0.7188498402555911, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1546, + "step": 900 + }, + { + "epoch": 0.7196485623003195, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1609, + "step": 901 + }, + { + "epoch": 0.7204472843450479, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1621, + "step": 902 + }, + { + "epoch": 0.7212460063897763, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1647, + "step": 903 + }, + { + "epoch": 0.7220447284345048, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1694, + "step": 904 + }, + { + "epoch": 0.7228434504792333, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1662, + "step": 905 + }, + { + "epoch": 0.7236421725239617, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1665, + "step": 906 + }, + { + "epoch": 0.7244408945686901, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1479, + "step": 907 + }, + { + "epoch": 0.7252396166134185, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1708, + "step": 908 + }, + { + "epoch": 0.7260383386581469, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1615, + "step": 909 + }, + { + "epoch": 0.7268370607028753, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1614, + "step": 910 + }, + { + "epoch": 0.7276357827476039, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1677, + "step": 911 + }, + { + "epoch": 0.7284345047923323, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1668, + "step": 912 + }, + { + "epoch": 0.7292332268370607, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.144, + "step": 913 + }, + { + "epoch": 0.7300319488817891, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1591, + "step": 914 + }, + { + "epoch": 0.7308306709265175, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1701, + "step": 915 + }, + { + "epoch": 0.731629392971246, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1747, + "step": 916 + }, + { + "epoch": 0.7324281150159745, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1688, + "step": 917 + }, + { + "epoch": 0.7332268370607029, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1534, + "step": 918 + }, + { + "epoch": 0.7340255591054313, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1637, + "step": 919 + }, + { + "epoch": 0.7348242811501597, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1677, + "step": 920 + }, + { + "epoch": 0.7356230031948882, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1661, + "step": 921 + }, + { + "epoch": 0.7364217252396166, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1606, + "step": 922 + }, + { + "epoch": 0.737220447284345, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.161, + "step": 923 + }, + { + "epoch": 0.7380191693290735, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1637, + "step": 924 + }, + { + "epoch": 0.7388178913738019, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1554, + "step": 925 + }, + { + "epoch": 0.7396166134185304, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1647, + "step": 926 + }, + { + "epoch": 0.7404153354632588, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1615, + "step": 927 + }, + { + "epoch": 0.7412140575079872, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1683, + "step": 928 + }, + { + "epoch": 0.7420127795527156, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.176, + "step": 929 + }, + { + "epoch": 0.7428115015974441, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1635, + "step": 930 + }, + { + "epoch": 0.7436102236421726, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.169, + "step": 931 + }, + { + "epoch": 0.744408945686901, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1667, + "step": 932 + }, + { + "epoch": 0.7452076677316294, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1647, + "step": 933 + }, + { + "epoch": 0.7460063897763578, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1668, + "step": 934 + }, + { + "epoch": 0.7468051118210862, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1565, + "step": 935 + }, + { + "epoch": 0.7476038338658147, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1522, + "step": 936 + }, + { + "epoch": 0.7484025559105432, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1623, + "step": 937 + }, + { + "epoch": 0.7492012779552716, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.162, + "step": 938 + }, + { + "epoch": 0.75, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.16, + "step": 939 + }, + { + "epoch": 0.7507987220447284, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1655, + "step": 940 + }, + { + "epoch": 0.7515974440894568, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1534, + "step": 941 + }, + { + "epoch": 0.7523961661341853, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.164, + "step": 942 + }, + { + "epoch": 0.7531948881789138, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1683, + "step": 943 + }, + { + "epoch": 0.7539936102236422, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1675, + "step": 944 + }, + { + "epoch": 0.7547923322683706, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1642, + "step": 945 + }, + { + "epoch": 0.755591054313099, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1644, + "step": 946 + }, + { + "epoch": 0.7563897763578274, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1648, + "step": 947 + }, + { + "epoch": 0.7571884984025559, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1626, + "step": 948 + }, + { + "epoch": 0.7579872204472844, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1492, + "step": 949 + }, + { + "epoch": 0.7587859424920128, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1612, + "step": 950 + }, + { + "epoch": 0.7595846645367412, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.165, + "step": 951 + }, + { + "epoch": 0.7603833865814696, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1571, + "step": 952 + }, + { + "epoch": 0.7611821086261981, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1537, + "step": 953 + }, + { + "epoch": 0.7619808306709265, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1624, + "step": 954 + }, + { + "epoch": 0.762779552715655, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1588, + "step": 955 + }, + { + "epoch": 0.7635782747603834, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1635, + "step": 956 + }, + { + "epoch": 0.7643769968051118, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1641, + "step": 957 + }, + { + "epoch": 0.7651757188498403, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1668, + "step": 958 + }, + { + "epoch": 0.7659744408945687, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1594, + "step": 959 + }, + { + "epoch": 0.7667731629392971, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1579, + "step": 960 + }, + { + "epoch": 0.7675718849840255, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1586, + "step": 961 + }, + { + "epoch": 0.768370607028754, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1653, + "step": 962 + }, + { + "epoch": 0.7691693290734825, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1591, + "step": 963 + }, + { + "epoch": 0.7699680511182109, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.164, + "step": 964 + }, + { + "epoch": 0.7707667731629393, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1613, + "step": 965 + }, + { + "epoch": 0.7715654952076677, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1623, + "step": 966 + }, + { + "epoch": 0.7723642172523961, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1616, + "step": 967 + }, + { + "epoch": 0.7731629392971247, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1646, + "step": 968 + }, + { + "epoch": 0.7739616613418531, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1705, + "step": 969 + }, + { + "epoch": 0.7747603833865815, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1544, + "step": 970 + }, + { + "epoch": 0.7755591054313099, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1567, + "step": 971 + }, + { + "epoch": 0.7763578274760383, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1533, + "step": 972 + }, + { + "epoch": 0.7771565495207667, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1585, + "step": 973 + }, + { + "epoch": 0.7779552715654952, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1529, + "step": 974 + }, + { + "epoch": 0.7787539936102237, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1589, + "step": 975 + }, + { + "epoch": 0.7795527156549521, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1427, + "step": 976 + }, + { + "epoch": 0.7803514376996805, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1583, + "step": 977 + }, + { + "epoch": 0.7811501597444089, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1651, + "step": 978 + }, + { + "epoch": 0.7819488817891374, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1581, + "step": 979 + }, + { + "epoch": 0.7827476038338658, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1512, + "step": 980 + }, + { + "epoch": 0.7835463258785943, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1566, + "step": 981 + }, + { + "epoch": 0.7843450479233227, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1662, + "step": 982 + }, + { + "epoch": 0.7851437699680511, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1697, + "step": 983 + }, + { + "epoch": 0.7859424920127795, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1514, + "step": 984 + }, + { + "epoch": 0.786741214057508, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1474, + "step": 985 + }, + { + "epoch": 0.7875399361022364, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1613, + "step": 986 + }, + { + "epoch": 0.7883386581469649, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1613, + "step": 987 + }, + { + "epoch": 0.7891373801916933, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.148, + "step": 988 + }, + { + "epoch": 0.7899361022364217, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1611, + "step": 989 + }, + { + "epoch": 0.7907348242811502, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.155, + "step": 990 + }, + { + "epoch": 0.7915335463258786, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1569, + "step": 991 + }, + { + "epoch": 0.792332268370607, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1525, + "step": 992 + }, + { + "epoch": 0.7931309904153354, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1595, + "step": 993 + }, + { + "epoch": 0.7939297124600639, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1699, + "step": 994 + }, + { + "epoch": 0.7947284345047924, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1555, + "step": 995 + }, + { + "epoch": 0.7955271565495208, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1611, + "step": 996 + }, + { + "epoch": 0.7963258785942492, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1619, + "step": 997 + }, + { + "epoch": 0.7971246006389776, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1572, + "step": 998 + }, + { + "epoch": 0.797923322683706, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1603, + "step": 999 + }, + { + "epoch": 0.7987220447284346, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1637, + "step": 1000 + }, + { + "epoch": 0.799520766773163, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1533, + "step": 1001 + }, + { + "epoch": 0.8003194888178914, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1568, + "step": 1002 + }, + { + "epoch": 0.8011182108626198, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1618, + "step": 1003 + }, + { + "epoch": 0.8019169329073482, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1495, + "step": 1004 + }, + { + "epoch": 0.8027156549520766, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1629, + "step": 1005 + }, + { + "epoch": 0.8035143769968051, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1617, + "step": 1006 + }, + { + "epoch": 0.8043130990415336, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1534, + "step": 1007 + }, + { + "epoch": 0.805111821086262, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1692, + "step": 1008 + }, + { + "epoch": 0.8059105431309904, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1606, + "step": 1009 + }, + { + "epoch": 0.8067092651757188, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1501, + "step": 1010 + }, + { + "epoch": 0.8075079872204473, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1573, + "step": 1011 + }, + { + "epoch": 0.8083067092651757, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1545, + "step": 1012 + }, + { + "epoch": 0.8091054313099042, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1591, + "step": 1013 + }, + { + "epoch": 0.8099041533546326, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.161, + "step": 1014 + }, + { + "epoch": 0.810702875399361, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.153, + "step": 1015 + }, + { + "epoch": 0.8115015974440895, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1625, + "step": 1016 + }, + { + "epoch": 0.8123003194888179, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1643, + "step": 1017 + }, + { + "epoch": 0.8130990415335463, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1492, + "step": 1018 + }, + { + "epoch": 0.8138977635782748, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1525, + "step": 1019 + }, + { + "epoch": 0.8146964856230032, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1566, + "step": 1020 + }, + { + "epoch": 0.8154952076677316, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1608, + "step": 1021 + }, + { + "epoch": 0.8162939297124601, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1445, + "step": 1022 + }, + { + "epoch": 0.8170926517571885, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1628, + "step": 1023 + }, + { + "epoch": 0.8178913738019169, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1587, + "step": 1024 + }, + { + "epoch": 0.8186900958466453, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1659, + "step": 1025 + }, + { + "epoch": 0.8194888178913738, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1579, + "step": 1026 + }, + { + "epoch": 0.8202875399361023, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1517, + "step": 1027 + }, + { + "epoch": 0.8210862619808307, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1623, + "step": 1028 + }, + { + "epoch": 0.8218849840255591, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.1615, + "step": 1029 + }, + { + "epoch": 0.8226837060702875, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1538, + "step": 1030 + }, + { + "epoch": 0.8234824281150159, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1495, + "step": 1031 + }, + { + "epoch": 0.8242811501597445, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.16, + "step": 1032 + }, + { + "epoch": 0.8250798722044729, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1604, + "step": 1033 + }, + { + "epoch": 0.8258785942492013, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1527, + "step": 1034 + }, + { + "epoch": 0.8266773162939297, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1619, + "step": 1035 + }, + { + "epoch": 0.8274760383386581, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1575, + "step": 1036 + }, + { + "epoch": 0.8282747603833865, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1709, + "step": 1037 + }, + { + "epoch": 0.829073482428115, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1604, + "step": 1038 + }, + { + "epoch": 0.8298722044728435, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1586, + "step": 1039 + }, + { + "epoch": 0.8306709265175719, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1582, + "step": 1040 + }, + { + "epoch": 0.8314696485623003, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1519, + "step": 1041 + }, + { + "epoch": 0.8322683706070287, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1554, + "step": 1042 + }, + { + "epoch": 0.8330670926517572, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1541, + "step": 1043 + }, + { + "epoch": 0.8338658146964856, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1563, + "step": 1044 + }, + { + "epoch": 0.8346645367412141, + "grad_norm": 0.39453125, + "learning_rate": 0.0001, + "loss": 1.1676, + "step": 1045 + }, + { + "epoch": 0.8354632587859425, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1573, + "step": 1046 + }, + { + "epoch": 0.8362619808306709, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1601, + "step": 1047 + }, + { + "epoch": 0.8370607028753994, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1535, + "step": 1048 + }, + { + "epoch": 0.8378594249201278, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1556, + "step": 1049 + }, + { + "epoch": 0.8386581469648562, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1647, + "step": 1050 + }, + { + "epoch": 0.8394568690095847, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.163, + "step": 1051 + }, + { + "epoch": 0.8402555910543131, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1524, + "step": 1052 + }, + { + "epoch": 0.8410543130990416, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1559, + "step": 1053 + }, + { + "epoch": 0.84185303514377, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1601, + "step": 1054 + }, + { + "epoch": 0.8426517571884984, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1494, + "step": 1055 + }, + { + "epoch": 0.8434504792332268, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1549, + "step": 1056 + }, + { + "epoch": 0.8442492012779552, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1526, + "step": 1057 + }, + { + "epoch": 0.8450479233226837, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1611, + "step": 1058 + }, + { + "epoch": 0.8458466453674122, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1598, + "step": 1059 + }, + { + "epoch": 0.8466453674121406, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1511, + "step": 1060 + }, + { + "epoch": 0.847444089456869, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1651, + "step": 1061 + }, + { + "epoch": 0.8482428115015974, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1601, + "step": 1062 + }, + { + "epoch": 0.8490415335463258, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1435, + "step": 1063 + }, + { + "epoch": 0.8498402555910544, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1547, + "step": 1064 + }, + { + "epoch": 0.8506389776357828, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1523, + "step": 1065 + }, + { + "epoch": 0.8514376996805112, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1551, + "step": 1066 + }, + { + "epoch": 0.8522364217252396, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1582, + "step": 1067 + }, + { + "epoch": 0.853035143769968, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1551, + "step": 1068 + }, + { + "epoch": 0.8538338658146964, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1506, + "step": 1069 + }, + { + "epoch": 0.854632587859425, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1606, + "step": 1070 + }, + { + "epoch": 0.8554313099041534, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1596, + "step": 1071 + }, + { + "epoch": 0.8562300319488818, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1631, + "step": 1072 + }, + { + "epoch": 0.8570287539936102, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.162, + "step": 1073 + }, + { + "epoch": 0.8578274760383386, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1586, + "step": 1074 + }, + { + "epoch": 0.8586261980830671, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1627, + "step": 1075 + }, + { + "epoch": 0.8594249201277955, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1508, + "step": 1076 + }, + { + "epoch": 0.860223642172524, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1713, + "step": 1077 + }, + { + "epoch": 0.8610223642172524, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1585, + "step": 1078 + }, + { + "epoch": 0.8618210862619808, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1345, + "step": 1079 + }, + { + "epoch": 0.8626198083067093, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1538, + "step": 1080 + }, + { + "epoch": 0.8634185303514377, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.153, + "step": 1081 + }, + { + "epoch": 0.8642172523961661, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1502, + "step": 1082 + }, + { + "epoch": 0.8650159744408946, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1614, + "step": 1083 + }, + { + "epoch": 0.865814696485623, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1604, + "step": 1084 + }, + { + "epoch": 0.8666134185303515, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1563, + "step": 1085 + }, + { + "epoch": 0.8674121405750799, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1498, + "step": 1086 + }, + { + "epoch": 0.8682108626198083, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1613, + "step": 1087 + }, + { + "epoch": 0.8690095846645367, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1592, + "step": 1088 + }, + { + "epoch": 0.8698083067092651, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1491, + "step": 1089 + }, + { + "epoch": 0.8706070287539937, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.152, + "step": 1090 + }, + { + "epoch": 0.8714057507987221, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1518, + "step": 1091 + }, + { + "epoch": 0.8722044728434505, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.154, + "step": 1092 + }, + { + "epoch": 0.8730031948881789, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1539, + "step": 1093 + }, + { + "epoch": 0.8738019169329073, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1524, + "step": 1094 + }, + { + "epoch": 0.8746006389776357, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1564, + "step": 1095 + }, + { + "epoch": 0.8753993610223643, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1572, + "step": 1096 + }, + { + "epoch": 0.8761980830670927, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1419, + "step": 1097 + }, + { + "epoch": 0.8769968051118211, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.16, + "step": 1098 + }, + { + "epoch": 0.8777955271565495, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1616, + "step": 1099 + }, + { + "epoch": 0.8785942492012779, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1626, + "step": 1100 + }, + { + "epoch": 0.8793929712460063, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1791, + "step": 1101 + }, + { + "epoch": 0.8801916932907349, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.154, + "step": 1102 + }, + { + "epoch": 0.8809904153354633, + "grad_norm": 0.345703125, + "learning_rate": 0.0001, + "loss": 1.1543, + "step": 1103 + }, + { + "epoch": 0.8817891373801917, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1577, + "step": 1104 + }, + { + "epoch": 0.8825878594249201, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.151, + "step": 1105 + }, + { + "epoch": 0.8833865814696485, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1585, + "step": 1106 + }, + { + "epoch": 0.884185303514377, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1497, + "step": 1107 + }, + { + "epoch": 0.8849840255591054, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1598, + "step": 1108 + }, + { + "epoch": 0.8857827476038339, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 1109 + }, + { + "epoch": 0.8865814696485623, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1645, + "step": 1110 + }, + { + "epoch": 0.8873801916932907, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1593, + "step": 1111 + }, + { + "epoch": 0.8881789137380192, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1522, + "step": 1112 + }, + { + "epoch": 0.8889776357827476, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1609, + "step": 1113 + }, + { + "epoch": 0.889776357827476, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1607, + "step": 1114 + }, + { + "epoch": 0.8905750798722045, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1536, + "step": 1115 + }, + { + "epoch": 0.8913738019169329, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1562, + "step": 1116 + }, + { + "epoch": 0.8921725239616614, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1519, + "step": 1117 + }, + { + "epoch": 0.8929712460063898, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1635, + "step": 1118 + }, + { + "epoch": 0.8937699680511182, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1553, + "step": 1119 + }, + { + "epoch": 0.8945686900958466, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.152, + "step": 1120 + }, + { + "epoch": 0.895367412140575, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1656, + "step": 1121 + }, + { + "epoch": 0.8961661341853036, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1552, + "step": 1122 + }, + { + "epoch": 0.896964856230032, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1563, + "step": 1123 + }, + { + "epoch": 0.8977635782747604, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.152, + "step": 1124 + }, + { + "epoch": 0.8985623003194888, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1566, + "step": 1125 + }, + { + "epoch": 0.8993610223642172, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1578, + "step": 1126 + }, + { + "epoch": 0.9001597444089456, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1502, + "step": 1127 + }, + { + "epoch": 0.9009584664536742, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1579, + "step": 1128 + }, + { + "epoch": 0.9017571884984026, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1572, + "step": 1129 + }, + { + "epoch": 0.902555910543131, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1515, + "step": 1130 + }, + { + "epoch": 0.9033546325878594, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1481, + "step": 1131 + }, + { + "epoch": 0.9041533546325878, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1592, + "step": 1132 + }, + { + "epoch": 0.9049520766773163, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1446, + "step": 1133 + }, + { + "epoch": 0.9057507987220448, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1564, + "step": 1134 + }, + { + "epoch": 0.9065495207667732, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1617, + "step": 1135 + }, + { + "epoch": 0.9073482428115016, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1562, + "step": 1136 + }, + { + "epoch": 0.90814696485623, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1612, + "step": 1137 + }, + { + "epoch": 0.9089456869009584, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1582, + "step": 1138 + }, + { + "epoch": 0.9097444089456869, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1591, + "step": 1139 + }, + { + "epoch": 0.9105431309904153, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1628, + "step": 1140 + }, + { + "epoch": 0.9113418530351438, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1563, + "step": 1141 + }, + { + "epoch": 0.9121405750798722, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.157, + "step": 1142 + }, + { + "epoch": 0.9129392971246006, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1554, + "step": 1143 + }, + { + "epoch": 0.9137380191693291, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1483, + "step": 1144 + }, + { + "epoch": 0.9145367412140575, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1416, + "step": 1145 + }, + { + "epoch": 0.9153354632587859, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1501, + "step": 1146 + }, + { + "epoch": 0.9161341853035144, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1448, + "step": 1147 + }, + { + "epoch": 0.9169329073482428, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1522, + "step": 1148 + }, + { + "epoch": 0.9177316293929713, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1575, + "step": 1149 + }, + { + "epoch": 0.9185303514376997, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1494, + "step": 1150 + }, + { + "epoch": 0.9193290734824281, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1638, + "step": 1151 + }, + { + "epoch": 0.9201277955271565, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1604, + "step": 1152 + }, + { + "epoch": 0.920926517571885, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1573, + "step": 1153 + }, + { + "epoch": 0.9217252396166135, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1476, + "step": 1154 + }, + { + "epoch": 0.9225239616613419, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1516, + "step": 1155 + }, + { + "epoch": 0.9233226837060703, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1559, + "step": 1156 + }, + { + "epoch": 0.9241214057507987, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1619, + "step": 1157 + }, + { + "epoch": 0.9249201277955271, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1577, + "step": 1158 + }, + { + "epoch": 0.9257188498402555, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1413, + "step": 1159 + }, + { + "epoch": 0.9265175718849841, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1556, + "step": 1160 + }, + { + "epoch": 0.9273162939297125, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1533, + "step": 1161 + }, + { + "epoch": 0.9281150159744409, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1576, + "step": 1162 + }, + { + "epoch": 0.9289137380191693, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1534, + "step": 1163 + }, + { + "epoch": 0.9297124600638977, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.153, + "step": 1164 + }, + { + "epoch": 0.9305111821086262, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1534, + "step": 1165 + }, + { + "epoch": 0.9313099041533547, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1609, + "step": 1166 + }, + { + "epoch": 0.9321086261980831, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1571, + "step": 1167 + }, + { + "epoch": 0.9329073482428115, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1597, + "step": 1168 + }, + { + "epoch": 0.9337060702875399, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.152, + "step": 1169 + }, + { + "epoch": 0.9345047923322684, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1564, + "step": 1170 + }, + { + "epoch": 0.9353035143769968, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1518, + "step": 1171 + }, + { + "epoch": 0.9361022364217252, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1612, + "step": 1172 + }, + { + "epoch": 0.9369009584664537, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1537, + "step": 1173 + }, + { + "epoch": 0.9376996805111821, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1542, + "step": 1174 + }, + { + "epoch": 0.9384984025559105, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1545, + "step": 1175 + }, + { + "epoch": 0.939297124600639, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.16, + "step": 1176 + }, + { + "epoch": 0.9400958466453674, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1589, + "step": 1177 + }, + { + "epoch": 0.9408945686900958, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1455, + "step": 1178 + }, + { + "epoch": 0.9416932907348243, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1536, + "step": 1179 + }, + { + "epoch": 0.9424920127795527, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1592, + "step": 1180 + }, + { + "epoch": 0.9432907348242812, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1503, + "step": 1181 + }, + { + "epoch": 0.9440894568690096, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1528, + "step": 1182 + }, + { + "epoch": 0.944888178913738, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1683, + "step": 1183 + }, + { + "epoch": 0.9456869009584664, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1457, + "step": 1184 + }, + { + "epoch": 0.9464856230031949, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1602, + "step": 1185 + }, + { + "epoch": 0.9472843450479234, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.155, + "step": 1186 + }, + { + "epoch": 0.9480830670926518, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1547, + "step": 1187 + }, + { + "epoch": 0.9488817891373802, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.152, + "step": 1188 + }, + { + "epoch": 0.9496805111821086, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1452, + "step": 1189 + }, + { + "epoch": 0.950479233226837, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1482, + "step": 1190 + }, + { + "epoch": 0.9512779552715654, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1523, + "step": 1191 + }, + { + "epoch": 0.952076677316294, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1561, + "step": 1192 + }, + { + "epoch": 0.9528753993610224, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 1193 + }, + { + "epoch": 0.9536741214057508, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1557, + "step": 1194 + }, + { + "epoch": 0.9544728434504792, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1463, + "step": 1195 + }, + { + "epoch": 0.9552715654952076, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1565, + "step": 1196 + }, + { + "epoch": 0.9560702875399361, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1585, + "step": 1197 + }, + { + "epoch": 0.9568690095846646, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1508, + "step": 1198 + }, + { + "epoch": 0.957667731629393, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1561, + "step": 1199 + }, + { + "epoch": 0.9584664536741214, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1595, + "step": 1200 + }, + { + "epoch": 0.9592651757188498, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1614, + "step": 1201 + }, + { + "epoch": 0.9600638977635783, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1549, + "step": 1202 + }, + { + "epoch": 0.9608626198083067, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1632, + "step": 1203 + }, + { + "epoch": 0.9616613418530351, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1469, + "step": 1204 + }, + { + "epoch": 0.9624600638977636, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1532, + "step": 1205 + }, + { + "epoch": 0.963258785942492, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.1563, + "step": 1206 + }, + { + "epoch": 0.9640575079872205, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.147, + "step": 1207 + }, + { + "epoch": 0.9648562300319489, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1511, + "step": 1208 + }, + { + "epoch": 0.9656549520766773, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.156, + "step": 1209 + }, + { + "epoch": 0.9664536741214057, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1545, + "step": 1210 + }, + { + "epoch": 0.9672523961661342, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1552, + "step": 1211 + }, + { + "epoch": 0.9680511182108626, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1553, + "step": 1212 + }, + { + "epoch": 0.9688498402555911, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1551, + "step": 1213 + }, + { + "epoch": 0.9696485623003195, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1558, + "step": 1214 + }, + { + "epoch": 0.9704472843450479, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1586, + "step": 1215 + }, + { + "epoch": 0.9712460063897763, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1492, + "step": 1216 + }, + { + "epoch": 0.9720447284345048, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1516, + "step": 1217 + }, + { + "epoch": 0.9728434504792333, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.159, + "step": 1218 + }, + { + "epoch": 0.9736421725239617, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 1219 + }, + { + "epoch": 0.9744408945686901, + "grad_norm": 0.384765625, + "learning_rate": 0.0001, + "loss": 1.1542, + "step": 1220 + }, + { + "epoch": 0.9752396166134185, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1478, + "step": 1221 + }, + { + "epoch": 0.9760383386581469, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1558, + "step": 1222 + }, + { + "epoch": 0.9768370607028753, + "grad_norm": 0.404296875, + "learning_rate": 0.0001, + "loss": 1.1534, + "step": 1223 + }, + { + "epoch": 0.9776357827476039, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1522, + "step": 1224 + }, + { + "epoch": 0.9784345047923323, + "grad_norm": 0.421875, + "learning_rate": 0.0001, + "loss": 1.1497, + "step": 1225 + }, + { + "epoch": 0.9792332268370607, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1476, + "step": 1226 + }, + { + "epoch": 0.9800319488817891, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1487, + "step": 1227 + }, + { + "epoch": 0.9808306709265175, + "grad_norm": 0.5, + "learning_rate": 0.0001, + "loss": 1.15, + "step": 1228 + }, + { + "epoch": 0.981629392971246, + "grad_norm": 0.470703125, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 1229 + }, + { + "epoch": 0.9824281150159745, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1544, + "step": 1230 + }, + { + "epoch": 0.9832268370607029, + "grad_norm": 0.5, + "learning_rate": 0.0001, + "loss": 1.1587, + "step": 1231 + }, + { + "epoch": 0.9840255591054313, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1556, + "step": 1232 + }, + { + "epoch": 0.9848242811501597, + "grad_norm": 0.337890625, + "learning_rate": 0.0001, + "loss": 1.1534, + "step": 1233 + }, + { + "epoch": 0.9856230031948882, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.153, + "step": 1234 + }, + { + "epoch": 0.9864217252396166, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1447, + "step": 1235 + }, + { + "epoch": 0.987220447284345, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 1236 + }, + { + "epoch": 0.9880191693290735, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1573, + "step": 1237 + }, + { + "epoch": 0.9888178913738019, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1466, + "step": 1238 + }, + { + "epoch": 0.9896166134185304, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 1239 + }, + { + "epoch": 0.9904153354632588, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1557, + "step": 1240 + }, + { + "epoch": 0.9912140575079872, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1584, + "step": 1241 + }, + { + "epoch": 0.9920127795527156, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.147, + "step": 1242 + }, + { + "epoch": 0.9928115015974441, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1521, + "step": 1243 + }, + { + "epoch": 0.9936102236421726, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1391, + "step": 1244 + }, + { + "epoch": 0.994408945686901, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1442, + "step": 1245 + }, + { + "epoch": 0.9952076677316294, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1626, + "step": 1246 + }, + { + "epoch": 0.9960063897763578, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1565, + "step": 1247 + }, + { + "epoch": 0.9968051118210862, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1517, + "step": 1248 + }, + { + "epoch": 0.9976038338658147, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1507, + "step": 1249 + }, + { + "epoch": 0.9984025559105432, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1496, + "step": 1250 + }, + { + "epoch": 0.9992012779552716, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1466, + "step": 1251 + }, + { + "epoch": 1.0, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1452, + "step": 1252 + }, + { + "epoch": 1.0007987220447285, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1512, + "step": 1253 + }, + { + "epoch": 1.0015974440894568, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1491, + "step": 1254 + }, + { + "epoch": 1.0023961661341854, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.152, + "step": 1255 + }, + { + "epoch": 1.0031948881789137, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.157, + "step": 1256 + }, + { + "epoch": 1.0039936102236422, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1508, + "step": 1257 + }, + { + "epoch": 1.0047923322683705, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1495, + "step": 1258 + }, + { + "epoch": 1.005591054313099, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1532, + "step": 1259 + }, + { + "epoch": 1.0063897763578276, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1635, + "step": 1260 + }, + { + "epoch": 1.0071884984025559, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1433, + "step": 1261 + }, + { + "epoch": 1.0079872204472844, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.156, + "step": 1262 + }, + { + "epoch": 1.0087859424920127, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 1263 + }, + { + "epoch": 1.0095846645367412, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1437, + "step": 1264 + }, + { + "epoch": 1.0103833865814698, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1563, + "step": 1265 + }, + { + "epoch": 1.011182108626198, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1472, + "step": 1266 + }, + { + "epoch": 1.0119808306709266, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1497, + "step": 1267 + }, + { + "epoch": 1.012779552715655, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1596, + "step": 1268 + }, + { + "epoch": 1.0135782747603834, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1484, + "step": 1269 + }, + { + "epoch": 1.0143769968051117, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1485, + "step": 1270 + }, + { + "epoch": 1.0151757188498403, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1499, + "step": 1271 + }, + { + "epoch": 1.0159744408945688, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1535, + "step": 1272 + }, + { + "epoch": 1.016773162939297, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1516, + "step": 1273 + }, + { + "epoch": 1.0175718849840256, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 1274 + }, + { + "epoch": 1.018370607028754, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1558, + "step": 1275 + }, + { + "epoch": 1.0191693290734825, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1477, + "step": 1276 + }, + { + "epoch": 1.0199680511182108, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.144, + "step": 1277 + }, + { + "epoch": 1.0207667731629393, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1547, + "step": 1278 + }, + { + "epoch": 1.0215654952076678, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 1279 + }, + { + "epoch": 1.0223642172523961, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.156, + "step": 1280 + }, + { + "epoch": 1.0231629392971247, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.151, + "step": 1281 + }, + { + "epoch": 1.023961661341853, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1423, + "step": 1282 + }, + { + "epoch": 1.0247603833865815, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 1283 + }, + { + "epoch": 1.0255591054313098, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1512, + "step": 1284 + }, + { + "epoch": 1.0263578274760383, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1508, + "step": 1285 + }, + { + "epoch": 1.0271565495207668, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1504, + "step": 1286 + }, + { + "epoch": 1.0279552715654952, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1611, + "step": 1287 + }, + { + "epoch": 1.0287539936102237, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1508, + "step": 1288 + }, + { + "epoch": 1.029552715654952, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1475, + "step": 1289 + }, + { + "epoch": 1.0303514376996805, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 1290 + }, + { + "epoch": 1.031150159744409, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 1291 + }, + { + "epoch": 1.0319488817891374, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1487, + "step": 1292 + }, + { + "epoch": 1.0327476038338659, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.1459, + "step": 1293 + }, + { + "epoch": 1.0335463258785942, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.1478, + "step": 1294 + }, + { + "epoch": 1.0343450479233227, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1448, + "step": 1295 + }, + { + "epoch": 1.035143769968051, + "grad_norm": 0.5078125, + "learning_rate": 0.0001, + "loss": 1.1558, + "step": 1296 + }, + { + "epoch": 1.0359424920127795, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.147, + "step": 1297 + }, + { + "epoch": 1.036741214057508, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1501, + "step": 1298 + }, + { + "epoch": 1.0375399361022364, + "grad_norm": 0.455078125, + "learning_rate": 0.0001, + "loss": 1.1543, + "step": 1299 + }, + { + "epoch": 1.038338658146965, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1481, + "step": 1300 + }, + { + "epoch": 1.0391373801916932, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1453, + "step": 1301 + }, + { + "epoch": 1.0399361022364217, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 1302 + }, + { + "epoch": 1.04073482428115, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 1303 + }, + { + "epoch": 1.0415335463258786, + "grad_norm": 0.466796875, + "learning_rate": 0.0001, + "loss": 1.146, + "step": 1304 + }, + { + "epoch": 1.042332268370607, + "grad_norm": 0.46484375, + "learning_rate": 0.0001, + "loss": 1.1579, + "step": 1305 + }, + { + "epoch": 1.0431309904153354, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1505, + "step": 1306 + }, + { + "epoch": 1.043929712460064, + "grad_norm": 0.46484375, + "learning_rate": 0.0001, + "loss": 1.1501, + "step": 1307 + }, + { + "epoch": 1.0447284345047922, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1534, + "step": 1308 + }, + { + "epoch": 1.0455271565495208, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.155, + "step": 1309 + }, + { + "epoch": 1.0463258785942493, + "grad_norm": 0.369140625, + "learning_rate": 0.0001, + "loss": 1.1448, + "step": 1310 + }, + { + "epoch": 1.0471246006389776, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 1311 + }, + { + "epoch": 1.0479233226837061, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1482, + "step": 1312 + }, + { + "epoch": 1.0487220447284344, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1446, + "step": 1313 + }, + { + "epoch": 1.049520766773163, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1472, + "step": 1314 + }, + { + "epoch": 1.0503194888178913, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1436, + "step": 1315 + }, + { + "epoch": 1.0511182108626198, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1575, + "step": 1316 + }, + { + "epoch": 1.0519169329073483, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1571, + "step": 1317 + }, + { + "epoch": 1.0527156549520766, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1563, + "step": 1318 + }, + { + "epoch": 1.0535143769968052, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1548, + "step": 1319 + }, + { + "epoch": 1.0543130990415335, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1519, + "step": 1320 + }, + { + "epoch": 1.055111821086262, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1506, + "step": 1321 + }, + { + "epoch": 1.0559105431309903, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1396, + "step": 1322 + }, + { + "epoch": 1.0567092651757188, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1481, + "step": 1323 + }, + { + "epoch": 1.0575079872204474, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.151, + "step": 1324 + }, + { + "epoch": 1.0583067092651757, + "grad_norm": 0.44921875, + "learning_rate": 0.0001, + "loss": 1.1504, + "step": 1325 + }, + { + "epoch": 1.0591054313099042, + "grad_norm": 0.373046875, + "learning_rate": 0.0001, + "loss": 1.1486, + "step": 1326 + }, + { + "epoch": 1.0599041533546325, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1455, + "step": 1327 + }, + { + "epoch": 1.060702875399361, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1557, + "step": 1328 + }, + { + "epoch": 1.0615015974440896, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1562, + "step": 1329 + }, + { + "epoch": 1.0623003194888179, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1527, + "step": 1330 + }, + { + "epoch": 1.0630990415335464, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1467, + "step": 1331 + }, + { + "epoch": 1.0638977635782747, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1522, + "step": 1332 + }, + { + "epoch": 1.0646964856230032, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1443, + "step": 1333 + }, + { + "epoch": 1.0654952076677315, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1504, + "step": 1334 + }, + { + "epoch": 1.06629392971246, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1453, + "step": 1335 + }, + { + "epoch": 1.0670926517571886, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1504, + "step": 1336 + }, + { + "epoch": 1.067891373801917, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1508, + "step": 1337 + }, + { + "epoch": 1.0686900958466454, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1492, + "step": 1338 + }, + { + "epoch": 1.0694888178913737, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1467, + "step": 1339 + }, + { + "epoch": 1.0702875399361023, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1396, + "step": 1340 + }, + { + "epoch": 1.0710862619808306, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1489, + "step": 1341 + }, + { + "epoch": 1.071884984025559, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1499, + "step": 1342 + }, + { + "epoch": 1.0726837060702876, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1507, + "step": 1343 + }, + { + "epoch": 1.073482428115016, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1555, + "step": 1344 + }, + { + "epoch": 1.0742811501597445, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1301, + "step": 1345 + }, + { + "epoch": 1.0750798722044728, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1515, + "step": 1346 + }, + { + "epoch": 1.0758785942492013, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 1347 + }, + { + "epoch": 1.0766773162939298, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1555, + "step": 1348 + }, + { + "epoch": 1.0774760383386581, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1575, + "step": 1349 + }, + { + "epoch": 1.0782747603833867, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1451, + "step": 1350 + }, + { + "epoch": 1.079073482428115, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1528, + "step": 1351 + }, + { + "epoch": 1.0798722044728435, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1391, + "step": 1352 + }, + { + "epoch": 1.0806709265175718, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1528, + "step": 1353 + }, + { + "epoch": 1.0814696485623003, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1516, + "step": 1354 + }, + { + "epoch": 1.0822683706070289, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1486, + "step": 1355 + }, + { + "epoch": 1.0830670926517572, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1476, + "step": 1356 + }, + { + "epoch": 1.0838658146964857, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1505, + "step": 1357 + }, + { + "epoch": 1.084664536741214, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1537, + "step": 1358 + }, + { + "epoch": 1.0854632587859425, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1457, + "step": 1359 + }, + { + "epoch": 1.0862619808306708, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1427, + "step": 1360 + }, + { + "epoch": 1.0870607028753994, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1501, + "step": 1361 + }, + { + "epoch": 1.0878594249201279, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 1362 + }, + { + "epoch": 1.0886581469648562, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1463, + "step": 1363 + }, + { + "epoch": 1.0894568690095847, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1449, + "step": 1364 + }, + { + "epoch": 1.090255591054313, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 1365 + }, + { + "epoch": 1.0910543130990416, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1523, + "step": 1366 + }, + { + "epoch": 1.09185303514377, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1546, + "step": 1367 + }, + { + "epoch": 1.0926517571884984, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1488, + "step": 1368 + }, + { + "epoch": 1.093450479233227, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1578, + "step": 1369 + }, + { + "epoch": 1.0942492012779552, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1502, + "step": 1370 + }, + { + "epoch": 1.0950479233226837, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 1371 + }, + { + "epoch": 1.095846645367412, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1577, + "step": 1372 + }, + { + "epoch": 1.0966453674121406, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1476, + "step": 1373 + }, + { + "epoch": 1.097444089456869, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1439, + "step": 1374 + }, + { + "epoch": 1.0982428115015974, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1487, + "step": 1375 + }, + { + "epoch": 1.099041533546326, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1544, + "step": 1376 + }, + { + "epoch": 1.0998402555910542, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1526, + "step": 1377 + }, + { + "epoch": 1.1006389776357828, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1434, + "step": 1378 + }, + { + "epoch": 1.101437699680511, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1488, + "step": 1379 + }, + { + "epoch": 1.1022364217252396, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.157, + "step": 1380 + }, + { + "epoch": 1.1030351437699681, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1542, + "step": 1381 + }, + { + "epoch": 1.1038338658146964, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1475, + "step": 1382 + }, + { + "epoch": 1.104632587859425, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1462, + "step": 1383 + }, + { + "epoch": 1.1054313099041533, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1544, + "step": 1384 + }, + { + "epoch": 1.1062300319488818, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1425, + "step": 1385 + }, + { + "epoch": 1.1070287539936103, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1506, + "step": 1386 + }, + { + "epoch": 1.1078274760383386, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1495, + "step": 1387 + }, + { + "epoch": 1.1086261980830672, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1406, + "step": 1388 + }, + { + "epoch": 1.1094249201277955, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1472, + "step": 1389 + }, + { + "epoch": 1.110223642172524, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1455, + "step": 1390 + }, + { + "epoch": 1.1110223642172523, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1393, + "step": 1391 + }, + { + "epoch": 1.1118210862619808, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1448, + "step": 1392 + }, + { + "epoch": 1.1126198083067094, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1422, + "step": 1393 + }, + { + "epoch": 1.1134185303514377, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1506, + "step": 1394 + }, + { + "epoch": 1.1142172523961662, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1497, + "step": 1395 + }, + { + "epoch": 1.1150159744408945, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1546, + "step": 1396 + }, + { + "epoch": 1.115814696485623, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1518, + "step": 1397 + }, + { + "epoch": 1.1166134185303513, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1555, + "step": 1398 + }, + { + "epoch": 1.1174121405750799, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1423, + "step": 1399 + }, + { + "epoch": 1.1182108626198084, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1531, + "step": 1400 + }, + { + "epoch": 1.1190095846645367, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1506, + "step": 1401 + }, + { + "epoch": 1.1198083067092652, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.1461, + "step": 1402 + }, + { + "epoch": 1.1206070287539935, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1571, + "step": 1403 + }, + { + "epoch": 1.121405750798722, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1605, + "step": 1404 + }, + { + "epoch": 1.1222044728434506, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1462, + "step": 1405 + }, + { + "epoch": 1.123003194888179, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.142, + "step": 1406 + }, + { + "epoch": 1.1238019169329074, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1502, + "step": 1407 + }, + { + "epoch": 1.1246006389776357, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1488, + "step": 1408 + }, + { + "epoch": 1.1253993610223643, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.15, + "step": 1409 + }, + { + "epoch": 1.1261980830670926, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1385, + "step": 1410 + }, + { + "epoch": 1.126996805111821, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1607, + "step": 1411 + }, + { + "epoch": 1.1277955271565494, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1489, + "step": 1412 + }, + { + "epoch": 1.128594249201278, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1452, + "step": 1413 + }, + { + "epoch": 1.1293929712460065, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1535, + "step": 1414 + }, + { + "epoch": 1.1301916932907348, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1486, + "step": 1415 + }, + { + "epoch": 1.1309904153354633, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1597, + "step": 1416 + }, + { + "epoch": 1.1317891373801916, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1507, + "step": 1417 + }, + { + "epoch": 1.1325878594249201, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1433, + "step": 1418 + }, + { + "epoch": 1.1333865814696487, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 1419 + }, + { + "epoch": 1.134185303514377, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1452, + "step": 1420 + }, + { + "epoch": 1.1349840255591055, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1407, + "step": 1421 + }, + { + "epoch": 1.1357827476038338, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1437, + "step": 1422 + }, + { + "epoch": 1.1365814696485623, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1507, + "step": 1423 + }, + { + "epoch": 1.1373801916932909, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1468, + "step": 1424 + }, + { + "epoch": 1.1381789137380192, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 1425 + }, + { + "epoch": 1.1389776357827477, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1493, + "step": 1426 + }, + { + "epoch": 1.139776357827476, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1528, + "step": 1427 + }, + { + "epoch": 1.1405750798722045, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1366, + "step": 1428 + }, + { + "epoch": 1.1413738019169328, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1568, + "step": 1429 + }, + { + "epoch": 1.1421725239616614, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1542, + "step": 1430 + }, + { + "epoch": 1.1429712460063897, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1474, + "step": 1431 + }, + { + "epoch": 1.1437699680511182, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1479, + "step": 1432 + }, + { + "epoch": 1.1445686900958467, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 1433 + }, + { + "epoch": 1.145367412140575, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1411, + "step": 1434 + }, + { + "epoch": 1.1461661341853036, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1463, + "step": 1435 + }, + { + "epoch": 1.1469648562300319, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1518, + "step": 1436 + }, + { + "epoch": 1.1477635782747604, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1508, + "step": 1437 + }, + { + "epoch": 1.148562300319489, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1526, + "step": 1438 + }, + { + "epoch": 1.1493610223642172, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1541, + "step": 1439 + }, + { + "epoch": 1.1501597444089458, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1552, + "step": 1440 + }, + { + "epoch": 1.150958466453674, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1418, + "step": 1441 + }, + { + "epoch": 1.1517571884984026, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1451, + "step": 1442 + }, + { + "epoch": 1.1525559105431311, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1563, + "step": 1443 + }, + { + "epoch": 1.1533546325878594, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1549, + "step": 1444 + }, + { + "epoch": 1.154153354632588, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1404, + "step": 1445 + }, + { + "epoch": 1.1549520766773163, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1394, + "step": 1446 + }, + { + "epoch": 1.1557507987220448, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 1447 + }, + { + "epoch": 1.156549520766773, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1517, + "step": 1448 + }, + { + "epoch": 1.1573482428115016, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1486, + "step": 1449 + }, + { + "epoch": 1.15814696485623, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1435, + "step": 1450 + }, + { + "epoch": 1.1589456869009584, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1411, + "step": 1451 + }, + { + "epoch": 1.159744408945687, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 1452 + }, + { + "epoch": 1.1605431309904153, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.145, + "step": 1453 + }, + { + "epoch": 1.1613418530351438, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1433, + "step": 1454 + }, + { + "epoch": 1.1621405750798721, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1501, + "step": 1455 + }, + { + "epoch": 1.1629392971246006, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 1456 + }, + { + "epoch": 1.1637380191693292, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.148, + "step": 1457 + }, + { + "epoch": 1.1645367412140575, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1527, + "step": 1458 + }, + { + "epoch": 1.165335463258786, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1492, + "step": 1459 + }, + { + "epoch": 1.1661341853035143, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1572, + "step": 1460 + }, + { + "epoch": 1.1669329073482428, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1593, + "step": 1461 + }, + { + "epoch": 1.1677316293929714, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1479, + "step": 1462 + }, + { + "epoch": 1.1685303514376997, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1514, + "step": 1463 + }, + { + "epoch": 1.1693290734824282, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1431, + "step": 1464 + }, + { + "epoch": 1.1701277955271565, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1483, + "step": 1465 + }, + { + "epoch": 1.170926517571885, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1454, + "step": 1466 + }, + { + "epoch": 1.1717252396166133, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1419, + "step": 1467 + }, + { + "epoch": 1.1725239616613419, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1487, + "step": 1468 + }, + { + "epoch": 1.1733226837060702, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1464, + "step": 1469 + }, + { + "epoch": 1.1741214057507987, + "grad_norm": 0.42578125, + "learning_rate": 0.0001, + "loss": 1.1503, + "step": 1470 + }, + { + "epoch": 1.1749201277955272, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1487, + "step": 1471 + }, + { + "epoch": 1.1757188498402555, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1424, + "step": 1472 + }, + { + "epoch": 1.176517571884984, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 1473 + }, + { + "epoch": 1.1773162939297124, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1335, + "step": 1474 + }, + { + "epoch": 1.178115015974441, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1518, + "step": 1475 + }, + { + "epoch": 1.1789137380191694, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1515, + "step": 1476 + }, + { + "epoch": 1.1797124600638977, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1408, + "step": 1477 + }, + { + "epoch": 1.1805111821086263, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1434, + "step": 1478 + }, + { + "epoch": 1.1813099041533546, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.149, + "step": 1479 + }, + { + "epoch": 1.182108626198083, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1527, + "step": 1480 + }, + { + "epoch": 1.1829073482428114, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1451, + "step": 1481 + }, + { + "epoch": 1.18370607028754, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1457, + "step": 1482 + }, + { + "epoch": 1.1845047923322685, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1442, + "step": 1483 + }, + { + "epoch": 1.1853035143769968, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1484, + "step": 1484 + }, + { + "epoch": 1.1861022364217253, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1479, + "step": 1485 + }, + { + "epoch": 1.1869009584664536, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 1486 + }, + { + "epoch": 1.1876996805111821, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1355, + "step": 1487 + }, + { + "epoch": 1.1884984025559104, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 1488 + }, + { + "epoch": 1.189297124600639, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1533, + "step": 1489 + }, + { + "epoch": 1.1900958466453675, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1467, + "step": 1490 + }, + { + "epoch": 1.1908945686900958, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.152, + "step": 1491 + }, + { + "epoch": 1.1916932907348243, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1474, + "step": 1492 + }, + { + "epoch": 1.1924920127795526, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1443, + "step": 1493 + }, + { + "epoch": 1.1932907348242812, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1448, + "step": 1494 + }, + { + "epoch": 1.1940894568690097, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1401, + "step": 1495 + }, + { + "epoch": 1.194888178913738, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1439, + "step": 1496 + }, + { + "epoch": 1.1956869009584665, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1497, + "step": 1497 + }, + { + "epoch": 1.1964856230031948, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1435, + "step": 1498 + }, + { + "epoch": 1.1972843450479234, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 1499 + }, + { + "epoch": 1.1980830670926517, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1528, + "step": 1500 + }, + { + "epoch": 1.1988817891373802, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1493, + "step": 1501 + }, + { + "epoch": 1.1996805111821087, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1472, + "step": 1502 + }, + { + "epoch": 1.200479233226837, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.145, + "step": 1503 + }, + { + "epoch": 1.2012779552715656, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 1504 + }, + { + "epoch": 1.2020766773162939, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1455, + "step": 1505 + }, + { + "epoch": 1.2028753993610224, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.154, + "step": 1506 + }, + { + "epoch": 1.2036741214057507, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1451, + "step": 1507 + }, + { + "epoch": 1.2044728434504792, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1483, + "step": 1508 + }, + { + "epoch": 1.2052715654952078, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1454, + "step": 1509 + }, + { + "epoch": 1.206070287539936, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1477, + "step": 1510 + }, + { + "epoch": 1.2068690095846646, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1451, + "step": 1511 + }, + { + "epoch": 1.207667731629393, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1369, + "step": 1512 + }, + { + "epoch": 1.2084664536741214, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1432, + "step": 1513 + }, + { + "epoch": 1.20926517571885, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 1514 + }, + { + "epoch": 1.2100638977635783, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1482, + "step": 1515 + }, + { + "epoch": 1.2108626198083068, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1406, + "step": 1516 + }, + { + "epoch": 1.211661341853035, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1503, + "step": 1517 + }, + { + "epoch": 1.2124600638977636, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 1518 + }, + { + "epoch": 1.213258785942492, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.145, + "step": 1519 + }, + { + "epoch": 1.2140575079872205, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1391, + "step": 1520 + }, + { + "epoch": 1.2148562300319488, + "grad_norm": 0.59765625, + "learning_rate": 0.0001, + "loss": 1.1546, + "step": 1521 + }, + { + "epoch": 1.2156549520766773, + "grad_norm": 0.625, + "learning_rate": 0.0001, + "loss": 1.1369, + "step": 1522 + }, + { + "epoch": 1.2164536741214058, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.149, + "step": 1523 + }, + { + "epoch": 1.2172523961661341, + "grad_norm": 0.419921875, + "learning_rate": 0.0001, + "loss": 1.1548, + "step": 1524 + }, + { + "epoch": 1.2180511182108626, + "grad_norm": 0.84375, + "learning_rate": 0.0001, + "loss": 1.1476, + "step": 1525 + }, + { + "epoch": 1.218849840255591, + "grad_norm": 1.4296875, + "learning_rate": 0.0001, + "loss": 1.159, + "step": 1526 + }, + { + "epoch": 1.2196485623003195, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1467, + "step": 1527 + }, + { + "epoch": 1.220447284345048, + "grad_norm": 1.5546875, + "learning_rate": 0.0001, + "loss": 1.1656, + "step": 1528 + }, + { + "epoch": 1.2212460063897763, + "grad_norm": 1.296875, + "learning_rate": 0.0001, + "loss": 1.1551, + "step": 1529 + }, + { + "epoch": 1.2220447284345048, + "grad_norm": 0.59375, + "learning_rate": 0.0001, + "loss": 1.1456, + "step": 1530 + }, + { + "epoch": 1.2228434504792332, + "grad_norm": 0.54296875, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 1531 + }, + { + "epoch": 1.2236421725239617, + "grad_norm": 0.515625, + "learning_rate": 0.0001, + "loss": 1.1566, + "step": 1532 + }, + { + "epoch": 1.2244408945686902, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.1497, + "step": 1533 + }, + { + "epoch": 1.2252396166134185, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.1491, + "step": 1534 + }, + { + "epoch": 1.226038338658147, + "grad_norm": 0.40234375, + "learning_rate": 0.0001, + "loss": 1.1337, + "step": 1535 + }, + { + "epoch": 1.2268370607028753, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1542, + "step": 1536 + }, + { + "epoch": 1.2276357827476039, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1611, + "step": 1537 + }, + { + "epoch": 1.2284345047923322, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1397, + "step": 1538 + }, + { + "epoch": 1.2292332268370607, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1463, + "step": 1539 + }, + { + "epoch": 1.230031948881789, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 1540 + }, + { + "epoch": 1.2308306709265175, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 1541 + }, + { + "epoch": 1.231629392971246, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1406, + "step": 1542 + }, + { + "epoch": 1.2324281150159744, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1441, + "step": 1543 + }, + { + "epoch": 1.233226837060703, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1429, + "step": 1544 + }, + { + "epoch": 1.2340255591054312, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 1545 + }, + { + "epoch": 1.2348242811501597, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1466, + "step": 1546 + }, + { + "epoch": 1.2356230031948883, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 1547 + }, + { + "epoch": 1.2364217252396166, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1526, + "step": 1548 + }, + { + "epoch": 1.237220447284345, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1481, + "step": 1549 + }, + { + "epoch": 1.2380191693290734, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1492, + "step": 1550 + }, + { + "epoch": 1.238817891373802, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1481, + "step": 1551 + }, + { + "epoch": 1.2396166134185305, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1516, + "step": 1552 + }, + { + "epoch": 1.2404153354632588, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1489, + "step": 1553 + }, + { + "epoch": 1.2412140575079873, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1468, + "step": 1554 + }, + { + "epoch": 1.2420127795527156, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1488, + "step": 1555 + }, + { + "epoch": 1.2428115015974441, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1437, + "step": 1556 + }, + { + "epoch": 1.2436102236421724, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1407, + "step": 1557 + }, + { + "epoch": 1.244408945686901, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.149, + "step": 1558 + }, + { + "epoch": 1.2452076677316293, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.159, + "step": 1559 + }, + { + "epoch": 1.2460063897763578, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1412, + "step": 1560 + }, + { + "epoch": 1.2468051118210863, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1455, + "step": 1561 + }, + { + "epoch": 1.2476038338658146, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1392, + "step": 1562 + }, + { + "epoch": 1.2484025559105432, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1525, + "step": 1563 + }, + { + "epoch": 1.2492012779552715, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1546, + "step": 1564 + }, + { + "epoch": 1.25, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1477, + "step": 1565 + }, + { + "epoch": 1.2507987220447285, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1509, + "step": 1566 + }, + { + "epoch": 1.2515974440894568, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1576, + "step": 1567 + }, + { + "epoch": 1.2523961661341854, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 1568 + }, + { + "epoch": 1.2531948881789137, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1295, + "step": 1569 + }, + { + "epoch": 1.2539936102236422, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1439, + "step": 1570 + }, + { + "epoch": 1.2547923322683707, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1467, + "step": 1571 + }, + { + "epoch": 1.255591054313099, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1497, + "step": 1572 + }, + { + "epoch": 1.2563897763578276, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1433, + "step": 1573 + }, + { + "epoch": 1.2571884984025559, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1495, + "step": 1574 + }, + { + "epoch": 1.2579872204472844, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1401, + "step": 1575 + }, + { + "epoch": 1.2587859424920127, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1471, + "step": 1576 + }, + { + "epoch": 1.2595846645367412, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1374, + "step": 1577 + }, + { + "epoch": 1.2603833865814695, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 1578 + }, + { + "epoch": 1.261182108626198, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1456, + "step": 1579 + }, + { + "epoch": 1.2619808306709266, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.1428, + "step": 1580 + }, + { + "epoch": 1.262779552715655, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1408, + "step": 1581 + }, + { + "epoch": 1.2635782747603834, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1428, + "step": 1582 + }, + { + "epoch": 1.2643769968051117, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1363, + "step": 1583 + }, + { + "epoch": 1.2651757188498403, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1456, + "step": 1584 + }, + { + "epoch": 1.2659744408945688, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1458, + "step": 1585 + }, + { + "epoch": 1.266773162939297, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 1586 + }, + { + "epoch": 1.2675718849840256, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1464, + "step": 1587 + }, + { + "epoch": 1.268370607028754, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1411, + "step": 1588 + }, + { + "epoch": 1.2691693290734825, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1498, + "step": 1589 + }, + { + "epoch": 1.269968051118211, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1428, + "step": 1590 + }, + { + "epoch": 1.2707667731629393, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 1591 + }, + { + "epoch": 1.2715654952076676, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1472, + "step": 1592 + }, + { + "epoch": 1.2723642172523961, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1557, + "step": 1593 + }, + { + "epoch": 1.2731629392971247, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1412, + "step": 1594 + }, + { + "epoch": 1.273961661341853, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1441, + "step": 1595 + }, + { + "epoch": 1.2747603833865815, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.147, + "step": 1596 + }, + { + "epoch": 1.2755591054313098, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 1597 + }, + { + "epoch": 1.2763578274760383, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 1598 + }, + { + "epoch": 1.2771565495207668, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1437, + "step": 1599 + }, + { + "epoch": 1.2779552715654952, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1521, + "step": 1600 + }, + { + "epoch": 1.2787539936102237, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1447, + "step": 1601 + }, + { + "epoch": 1.279552715654952, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1422, + "step": 1602 + }, + { + "epoch": 1.2803514376996805, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.147, + "step": 1603 + }, + { + "epoch": 1.281150159744409, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1499, + "step": 1604 + }, + { + "epoch": 1.2819488817891374, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 1605 + }, + { + "epoch": 1.2827476038338659, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1466, + "step": 1606 + }, + { + "epoch": 1.2835463258785942, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.138, + "step": 1607 + }, + { + "epoch": 1.2843450479233227, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1419, + "step": 1608 + }, + { + "epoch": 1.2851437699680512, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.143, + "step": 1609 + }, + { + "epoch": 1.2859424920127795, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1465, + "step": 1610 + }, + { + "epoch": 1.2867412140575079, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1495, + "step": 1611 + }, + { + "epoch": 1.2875399361022364, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1346, + "step": 1612 + }, + { + "epoch": 1.288338658146965, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1487, + "step": 1613 + }, + { + "epoch": 1.2891373801916932, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.138, + "step": 1614 + }, + { + "epoch": 1.2899361022364217, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1428, + "step": 1615 + }, + { + "epoch": 1.29073482428115, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1391, + "step": 1616 + }, + { + "epoch": 1.2915335463258786, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1521, + "step": 1617 + }, + { + "epoch": 1.292332268370607, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 1618 + }, + { + "epoch": 1.2931309904153354, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1419, + "step": 1619 + }, + { + "epoch": 1.293929712460064, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1547, + "step": 1620 + }, + { + "epoch": 1.2947284345047922, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.146, + "step": 1621 + }, + { + "epoch": 1.2955271565495208, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 1622 + }, + { + "epoch": 1.2963258785942493, + "grad_norm": 0.359375, + "learning_rate": 0.0001, + "loss": 1.152, + "step": 1623 + }, + { + "epoch": 1.2971246006389776, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1465, + "step": 1624 + }, + { + "epoch": 1.2979233226837061, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1361, + "step": 1625 + }, + { + "epoch": 1.2987220447284344, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1357, + "step": 1626 + }, + { + "epoch": 1.299520766773163, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1416, + "step": 1627 + }, + { + "epoch": 1.3003194888178915, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 1628 + }, + { + "epoch": 1.3011182108626198, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1432, + "step": 1629 + }, + { + "epoch": 1.3019169329073481, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1447, + "step": 1630 + }, + { + "epoch": 1.3027156549520766, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 1631 + }, + { + "epoch": 1.3035143769968052, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1439, + "step": 1632 + }, + { + "epoch": 1.3043130990415335, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1405, + "step": 1633 + }, + { + "epoch": 1.305111821086262, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1406, + "step": 1634 + }, + { + "epoch": 1.3059105431309903, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1419, + "step": 1635 + }, + { + "epoch": 1.3067092651757188, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1398, + "step": 1636 + }, + { + "epoch": 1.3075079872204474, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1509, + "step": 1637 + }, + { + "epoch": 1.3083067092651757, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1391, + "step": 1638 + }, + { + "epoch": 1.3091054313099042, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1464, + "step": 1639 + }, + { + "epoch": 1.3099041533546325, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1486, + "step": 1640 + }, + { + "epoch": 1.310702875399361, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1455, + "step": 1641 + }, + { + "epoch": 1.3115015974440896, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.147, + "step": 1642 + }, + { + "epoch": 1.3123003194888179, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1464, + "step": 1643 + }, + { + "epoch": 1.3130990415335464, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1328, + "step": 1644 + }, + { + "epoch": 1.3138977635782747, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1395, + "step": 1645 + }, + { + "epoch": 1.3146964856230032, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1363, + "step": 1646 + }, + { + "epoch": 1.3154952076677318, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1508, + "step": 1647 + }, + { + "epoch": 1.31629392971246, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1377, + "step": 1648 + }, + { + "epoch": 1.3170926517571884, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.148, + "step": 1649 + }, + { + "epoch": 1.317891373801917, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1428, + "step": 1650 + }, + { + "epoch": 1.3186900958466454, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1445, + "step": 1651 + }, + { + "epoch": 1.3194888178913737, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1462, + "step": 1652 + }, + { + "epoch": 1.3202875399361023, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1481, + "step": 1653 + }, + { + "epoch": 1.3210862619808306, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1495, + "step": 1654 + }, + { + "epoch": 1.321884984025559, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1459, + "step": 1655 + }, + { + "epoch": 1.3226837060702876, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1453, + "step": 1656 + }, + { + "epoch": 1.323482428115016, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1346, + "step": 1657 + }, + { + "epoch": 1.3242811501597445, + "grad_norm": 0.408203125, + "learning_rate": 0.0001, + "loss": 1.1462, + "step": 1658 + }, + { + "epoch": 1.3250798722044728, + "grad_norm": 0.396484375, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 1659 + }, + { + "epoch": 1.3258785942492013, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1467, + "step": 1660 + }, + { + "epoch": 1.3266773162939298, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1373, + "step": 1661 + }, + { + "epoch": 1.3274760383386581, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 1662 + }, + { + "epoch": 1.3282747603833867, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1411, + "step": 1663 + }, + { + "epoch": 1.329073482428115, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1387, + "step": 1664 + }, + { + "epoch": 1.3298722044728435, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.151, + "step": 1665 + }, + { + "epoch": 1.330670926517572, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 1666 + }, + { + "epoch": 1.3314696485623003, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 1667 + }, + { + "epoch": 1.3322683706070286, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.145, + "step": 1668 + }, + { + "epoch": 1.3330670926517572, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1461, + "step": 1669 + }, + { + "epoch": 1.3338658146964857, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1425, + "step": 1670 + }, + { + "epoch": 1.334664536741214, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1363, + "step": 1671 + }, + { + "epoch": 1.3354632587859425, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1455, + "step": 1672 + }, + { + "epoch": 1.3362619808306708, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1419, + "step": 1673 + }, + { + "epoch": 1.3370607028753994, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 1674 + }, + { + "epoch": 1.3378594249201279, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1377, + "step": 1675 + }, + { + "epoch": 1.3386581469648562, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1461, + "step": 1676 + }, + { + "epoch": 1.3394568690095847, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1435, + "step": 1677 + }, + { + "epoch": 1.340255591054313, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1398, + "step": 1678 + }, + { + "epoch": 1.3410543130990416, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1374, + "step": 1679 + }, + { + "epoch": 1.34185303514377, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1432, + "step": 1680 + }, + { + "epoch": 1.3426517571884984, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1397, + "step": 1681 + }, + { + "epoch": 1.343450479233227, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1457, + "step": 1682 + }, + { + "epoch": 1.3442492012779552, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1368, + "step": 1683 + }, + { + "epoch": 1.3450479233226837, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1435, + "step": 1684 + }, + { + "epoch": 1.3458466453674123, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1481, + "step": 1685 + }, + { + "epoch": 1.3466453674121406, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1429, + "step": 1686 + }, + { + "epoch": 1.3474440894568689, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1433, + "step": 1687 + }, + { + "epoch": 1.3482428115015974, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1477, + "step": 1688 + }, + { + "epoch": 1.349041533546326, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 1689 + }, + { + "epoch": 1.3498402555910542, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1408, + "step": 1690 + }, + { + "epoch": 1.3506389776357828, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1381, + "step": 1691 + }, + { + "epoch": 1.351437699680511, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 1692 + }, + { + "epoch": 1.3522364217252396, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 1693 + }, + { + "epoch": 1.3530351437699681, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1466, + "step": 1694 + }, + { + "epoch": 1.3538338658146964, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1433, + "step": 1695 + }, + { + "epoch": 1.354632587859425, + "grad_norm": 0.384765625, + "learning_rate": 0.0001, + "loss": 1.1448, + "step": 1696 + }, + { + "epoch": 1.3554313099041533, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1467, + "step": 1697 + }, + { + "epoch": 1.3562300319488818, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1301, + "step": 1698 + }, + { + "epoch": 1.3570287539936103, + "grad_norm": 0.419921875, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 1699 + }, + { + "epoch": 1.3578274760383386, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1386, + "step": 1700 + }, + { + "epoch": 1.3586261980830672, + "grad_norm": 0.435546875, + "learning_rate": 0.0001, + "loss": 1.1471, + "step": 1701 + }, + { + "epoch": 1.3594249201277955, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1424, + "step": 1702 + }, + { + "epoch": 1.360223642172524, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1503, + "step": 1703 + }, + { + "epoch": 1.3610223642172525, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1502, + "step": 1704 + }, + { + "epoch": 1.3618210862619808, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1443, + "step": 1705 + }, + { + "epoch": 1.3626198083067091, + "grad_norm": 0.328125, + "learning_rate": 0.0001, + "loss": 1.1505, + "step": 1706 + }, + { + "epoch": 1.3634185303514377, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 1707 + }, + { + "epoch": 1.3642172523961662, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1459, + "step": 1708 + }, + { + "epoch": 1.3650159744408945, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1382, + "step": 1709 + }, + { + "epoch": 1.365814696485623, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1452, + "step": 1710 + }, + { + "epoch": 1.3666134185303513, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1599, + "step": 1711 + }, + { + "epoch": 1.3674121405750799, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1375, + "step": 1712 + }, + { + "epoch": 1.3682108626198084, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1449, + "step": 1713 + }, + { + "epoch": 1.3690095846645367, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1401, + "step": 1714 + }, + { + "epoch": 1.3698083067092652, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 1715 + }, + { + "epoch": 1.3706070287539935, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1503, + "step": 1716 + }, + { + "epoch": 1.371405750798722, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1512, + "step": 1717 + }, + { + "epoch": 1.3722044728434506, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1419, + "step": 1718 + }, + { + "epoch": 1.373003194888179, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 1719 + }, + { + "epoch": 1.3738019169329074, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1494, + "step": 1720 + }, + { + "epoch": 1.3746006389776357, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1481, + "step": 1721 + }, + { + "epoch": 1.3753993610223643, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1461, + "step": 1722 + }, + { + "epoch": 1.3761980830670926, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 1723 + }, + { + "epoch": 1.376996805111821, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1397, + "step": 1724 + }, + { + "epoch": 1.3777955271565494, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1415, + "step": 1725 + }, + { + "epoch": 1.378594249201278, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.138, + "step": 1726 + }, + { + "epoch": 1.3793929712460065, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1512, + "step": 1727 + }, + { + "epoch": 1.3801916932907348, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1525, + "step": 1728 + }, + { + "epoch": 1.3809904153354633, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.1486, + "step": 1729 + }, + { + "epoch": 1.3817891373801916, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1408, + "step": 1730 + }, + { + "epoch": 1.3825878594249201, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1382, + "step": 1731 + }, + { + "epoch": 1.3833865814696487, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 1732 + }, + { + "epoch": 1.384185303514377, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1522, + "step": 1733 + }, + { + "epoch": 1.3849840255591055, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 1734 + }, + { + "epoch": 1.3857827476038338, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 1735 + }, + { + "epoch": 1.3865814696485623, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1467, + "step": 1736 + }, + { + "epoch": 1.3873801916932909, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1415, + "step": 1737 + }, + { + "epoch": 1.3881789137380192, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1375, + "step": 1738 + }, + { + "epoch": 1.3889776357827475, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1478, + "step": 1739 + }, + { + "epoch": 1.389776357827476, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.149, + "step": 1740 + }, + { + "epoch": 1.3905750798722045, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1441, + "step": 1741 + }, + { + "epoch": 1.3913738019169328, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1418, + "step": 1742 + }, + { + "epoch": 1.3921725239616614, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.144, + "step": 1743 + }, + { + "epoch": 1.3929712460063897, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1498, + "step": 1744 + }, + { + "epoch": 1.3937699680511182, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 1745 + }, + { + "epoch": 1.3945686900958467, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1385, + "step": 1746 + }, + { + "epoch": 1.395367412140575, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1381, + "step": 1747 + }, + { + "epoch": 1.3961661341853036, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1487, + "step": 1748 + }, + { + "epoch": 1.3969648562300319, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1404, + "step": 1749 + }, + { + "epoch": 1.3977635782747604, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1443, + "step": 1750 + }, + { + "epoch": 1.398562300319489, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1463, + "step": 1751 + }, + { + "epoch": 1.3993610223642172, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1406, + "step": 1752 + }, + { + "epoch": 1.4001597444089458, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1455, + "step": 1753 + }, + { + "epoch": 1.400958466453674, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1522, + "step": 1754 + }, + { + "epoch": 1.4017571884984026, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 1755 + }, + { + "epoch": 1.4025559105431311, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 1756 + }, + { + "epoch": 1.4033546325878594, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 1757 + }, + { + "epoch": 1.4041533546325877, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1486, + "step": 1758 + }, + { + "epoch": 1.4049520766773163, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1477, + "step": 1759 + }, + { + "epoch": 1.4057507987220448, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1411, + "step": 1760 + }, + { + "epoch": 1.406549520766773, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1407, + "step": 1761 + }, + { + "epoch": 1.4073482428115016, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1404, + "step": 1762 + }, + { + "epoch": 1.40814696485623, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1405, + "step": 1763 + }, + { + "epoch": 1.4089456869009584, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1451, + "step": 1764 + }, + { + "epoch": 1.409744408945687, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.15, + "step": 1765 + }, + { + "epoch": 1.4105431309904153, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1373, + "step": 1766 + }, + { + "epoch": 1.4113418530351438, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1383, + "step": 1767 + }, + { + "epoch": 1.4121405750798721, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1378, + "step": 1768 + }, + { + "epoch": 1.4129392971246006, + "grad_norm": 0.361328125, + "learning_rate": 0.0001, + "loss": 1.1433, + "step": 1769 + }, + { + "epoch": 1.4137380191693292, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1459, + "step": 1770 + }, + { + "epoch": 1.4145367412140575, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1418, + "step": 1771 + }, + { + "epoch": 1.415335463258786, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1429, + "step": 1772 + }, + { + "epoch": 1.4161341853035143, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1441, + "step": 1773 + }, + { + "epoch": 1.4169329073482428, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1446, + "step": 1774 + }, + { + "epoch": 1.4177316293929714, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1387, + "step": 1775 + }, + { + "epoch": 1.4185303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 1776 + }, + { + "epoch": 1.419329073482428, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1401, + "step": 1777 + }, + { + "epoch": 1.4201277955271565, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 1778 + }, + { + "epoch": 1.420926517571885, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1404, + "step": 1779 + }, + { + "epoch": 1.4217252396166133, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1442, + "step": 1780 + }, + { + "epoch": 1.4225239616613419, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1502, + "step": 1781 + }, + { + "epoch": 1.4233226837060702, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 1782 + }, + { + "epoch": 1.4241214057507987, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1394, + "step": 1783 + }, + { + "epoch": 1.4249201277955272, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.142, + "step": 1784 + }, + { + "epoch": 1.4257188498402555, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 1785 + }, + { + "epoch": 1.426517571884984, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1368, + "step": 1786 + }, + { + "epoch": 1.4273162939297124, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.142, + "step": 1787 + }, + { + "epoch": 1.428115015974441, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 1788 + }, + { + "epoch": 1.4289137380191694, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 1789 + }, + { + "epoch": 1.4297124600638977, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1381, + "step": 1790 + }, + { + "epoch": 1.4305111821086263, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1485, + "step": 1791 + }, + { + "epoch": 1.4313099041533546, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 1792 + }, + { + "epoch": 1.432108626198083, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1398, + "step": 1793 + }, + { + "epoch": 1.4329073482428116, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 1794 + }, + { + "epoch": 1.43370607028754, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 1795 + }, + { + "epoch": 1.4345047923322682, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1464, + "step": 1796 + }, + { + "epoch": 1.4353035143769968, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1459, + "step": 1797 + }, + { + "epoch": 1.4361022364217253, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 1798 + }, + { + "epoch": 1.4369009584664536, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.1394, + "step": 1799 + }, + { + "epoch": 1.4376996805111821, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.152, + "step": 1800 + }, + { + "epoch": 1.4384984025559104, + "grad_norm": 0.443359375, + "learning_rate": 0.0001, + "loss": 1.1458, + "step": 1801 + }, + { + "epoch": 1.439297124600639, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1376, + "step": 1802 + }, + { + "epoch": 1.4400958466453675, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1518, + "step": 1803 + }, + { + "epoch": 1.4408945686900958, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.1431, + "step": 1804 + }, + { + "epoch": 1.4416932907348243, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 1805 + }, + { + "epoch": 1.4424920127795526, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1527, + "step": 1806 + }, + { + "epoch": 1.4432907348242812, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 1807 + }, + { + "epoch": 1.4440894568690097, + "grad_norm": 0.357421875, + "learning_rate": 0.0001, + "loss": 1.1487, + "step": 1808 + }, + { + "epoch": 1.444888178913738, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1425, + "step": 1809 + }, + { + "epoch": 1.4456869009584665, + "grad_norm": 0.34375, + "learning_rate": 0.0001, + "loss": 1.1435, + "step": 1810 + }, + { + "epoch": 1.4464856230031948, + "grad_norm": 0.357421875, + "learning_rate": 0.0001, + "loss": 1.1468, + "step": 1811 + }, + { + "epoch": 1.4472843450479234, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1311, + "step": 1812 + }, + { + "epoch": 1.4480830670926519, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1491, + "step": 1813 + }, + { + "epoch": 1.4488817891373802, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 1814 + }, + { + "epoch": 1.4496805111821085, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1496, + "step": 1815 + }, + { + "epoch": 1.450479233226837, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 1816 + }, + { + "epoch": 1.4512779552715656, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 1817 + }, + { + "epoch": 1.4520766773162939, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.149, + "step": 1818 + }, + { + "epoch": 1.4528753993610224, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1495, + "step": 1819 + }, + { + "epoch": 1.4536741214057507, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1349, + "step": 1820 + }, + { + "epoch": 1.4544728434504792, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1455, + "step": 1821 + }, + { + "epoch": 1.4552715654952078, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 1822 + }, + { + "epoch": 1.456070287539936, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1447, + "step": 1823 + }, + { + "epoch": 1.4568690095846646, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 1824 + }, + { + "epoch": 1.457667731629393, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1415, + "step": 1825 + }, + { + "epoch": 1.4584664536741214, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.135, + "step": 1826 + }, + { + "epoch": 1.45926517571885, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 1827 + }, + { + "epoch": 1.4600638977635783, + "grad_norm": 0.37109375, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 1828 + }, + { + "epoch": 1.4608626198083068, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 1829 + }, + { + "epoch": 1.461661341853035, + "grad_norm": 0.384765625, + "learning_rate": 0.0001, + "loss": 1.1393, + "step": 1830 + }, + { + "epoch": 1.4624600638977636, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 1831 + }, + { + "epoch": 1.4632587859424921, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 1832 + }, + { + "epoch": 1.4640575079872205, + "grad_norm": 0.4296875, + "learning_rate": 0.0001, + "loss": 1.1476, + "step": 1833 + }, + { + "epoch": 1.4648562300319488, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1456, + "step": 1834 + }, + { + "epoch": 1.4656549520766773, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 1835 + }, + { + "epoch": 1.4664536741214058, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 1836 + }, + { + "epoch": 1.4672523961661341, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 1837 + }, + { + "epoch": 1.4680511182108626, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.144, + "step": 1838 + }, + { + "epoch": 1.468849840255591, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1492, + "step": 1839 + }, + { + "epoch": 1.4696485623003195, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1482, + "step": 1840 + }, + { + "epoch": 1.470447284345048, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 1841 + }, + { + "epoch": 1.4712460063897763, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1563, + "step": 1842 + }, + { + "epoch": 1.4720447284345048, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1431, + "step": 1843 + }, + { + "epoch": 1.4728434504792332, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 1844 + }, + { + "epoch": 1.4736421725239617, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 1845 + }, + { + "epoch": 1.4744408945686902, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 1846 + }, + { + "epoch": 1.4752396166134185, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1432, + "step": 1847 + }, + { + "epoch": 1.476038338658147, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1376, + "step": 1848 + }, + { + "epoch": 1.4768370607028753, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1436, + "step": 1849 + }, + { + "epoch": 1.4776357827476039, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1434, + "step": 1850 + }, + { + "epoch": 1.4784345047923324, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1451, + "step": 1851 + }, + { + "epoch": 1.4792332268370607, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 1852 + }, + { + "epoch": 1.480031948881789, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1394, + "step": 1853 + }, + { + "epoch": 1.4808306709265175, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1548, + "step": 1854 + }, + { + "epoch": 1.481629392971246, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 1855 + }, + { + "epoch": 1.4824281150159744, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 1856 + }, + { + "epoch": 1.483226837060703, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1305, + "step": 1857 + }, + { + "epoch": 1.4840255591054312, + "grad_norm": 0.40625, + "learning_rate": 0.0001, + "loss": 1.1368, + "step": 1858 + }, + { + "epoch": 1.4848242811501597, + "grad_norm": 0.43359375, + "learning_rate": 0.0001, + "loss": 1.1393, + "step": 1859 + }, + { + "epoch": 1.4856230031948883, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 1860 + }, + { + "epoch": 1.4864217252396166, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 1861 + }, + { + "epoch": 1.487220447284345, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 1862 + }, + { + "epoch": 1.4880191693290734, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 1863 + }, + { + "epoch": 1.488817891373802, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1396, + "step": 1864 + }, + { + "epoch": 1.4896166134185305, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1355, + "step": 1865 + }, + { + "epoch": 1.4904153354632588, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 1866 + }, + { + "epoch": 1.4912140575079873, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 1867 + }, + { + "epoch": 1.4920127795527156, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 1868 + }, + { + "epoch": 1.4928115015974441, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 1869 + }, + { + "epoch": 1.4936102236421724, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 1870 + }, + { + "epoch": 1.494408945686901, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 1871 + }, + { + "epoch": 1.4952076677316293, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1423, + "step": 1872 + }, + { + "epoch": 1.4960063897763578, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.146, + "step": 1873 + }, + { + "epoch": 1.4968051118210863, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1457, + "step": 1874 + }, + { + "epoch": 1.4976038338658146, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 1875 + }, + { + "epoch": 1.4984025559105432, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 1876 + }, + { + "epoch": 1.4992012779552715, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 1877 + }, + { + "epoch": 1.5, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1425, + "step": 1878 + }, + { + "epoch": 1.5007987220447285, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1397, + "step": 1879 + }, + { + "epoch": 1.5015974440894568, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 1880 + }, + { + "epoch": 1.5023961661341851, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 1881 + }, + { + "epoch": 1.5031948881789137, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 1882 + }, + { + "epoch": 1.5039936102236422, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 1883 + }, + { + "epoch": 1.5047923322683707, + "grad_norm": 0.357421875, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 1884 + }, + { + "epoch": 1.505591054313099, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 1885 + }, + { + "epoch": 1.5063897763578273, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1408, + "step": 1886 + }, + { + "epoch": 1.5071884984025559, + "grad_norm": 0.39453125, + "learning_rate": 0.0001, + "loss": 1.1397, + "step": 1887 + }, + { + "epoch": 1.5079872204472844, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.143, + "step": 1888 + }, + { + "epoch": 1.508785942492013, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1384, + "step": 1889 + }, + { + "epoch": 1.5095846645367412, + "grad_norm": 0.375, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 1890 + }, + { + "epoch": 1.5103833865814695, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1478, + "step": 1891 + }, + { + "epoch": 1.511182108626198, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1375, + "step": 1892 + }, + { + "epoch": 1.5119808306709266, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 1893 + }, + { + "epoch": 1.5127795527156551, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 1894 + }, + { + "epoch": 1.5135782747603834, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1487, + "step": 1895 + }, + { + "epoch": 1.5143769968051117, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1439, + "step": 1896 + }, + { + "epoch": 1.5151757188498403, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 1897 + }, + { + "epoch": 1.5159744408945688, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1361, + "step": 1898 + }, + { + "epoch": 1.516773162939297, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1325, + "step": 1899 + }, + { + "epoch": 1.5175718849840254, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.141, + "step": 1900 + }, + { + "epoch": 1.518370607028754, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.135, + "step": 1901 + }, + { + "epoch": 1.5191693290734825, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.146, + "step": 1902 + }, + { + "epoch": 1.519968051118211, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1359, + "step": 1903 + }, + { + "epoch": 1.5207667731629393, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 1904 + }, + { + "epoch": 1.5215654952076676, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1435, + "step": 1905 + }, + { + "epoch": 1.5223642172523961, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 1906 + }, + { + "epoch": 1.5231629392971247, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1414, + "step": 1907 + }, + { + "epoch": 1.5239616613418532, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1375, + "step": 1908 + }, + { + "epoch": 1.5247603833865815, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1387, + "step": 1909 + }, + { + "epoch": 1.5255591054313098, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1399, + "step": 1910 + }, + { + "epoch": 1.5263578274760383, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1404, + "step": 1911 + }, + { + "epoch": 1.5271565495207668, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1492, + "step": 1912 + }, + { + "epoch": 1.5279552715654952, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 1913 + }, + { + "epoch": 1.5287539936102237, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 1914 + }, + { + "epoch": 1.529552715654952, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 1915 + }, + { + "epoch": 1.5303514376996805, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.134, + "step": 1916 + }, + { + "epoch": 1.531150159744409, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 1917 + }, + { + "epoch": 1.5319488817891374, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 1918 + }, + { + "epoch": 1.5327476038338657, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 1919 + }, + { + "epoch": 1.5335463258785942, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 1920 + }, + { + "epoch": 1.5343450479233227, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1448, + "step": 1921 + }, + { + "epoch": 1.5351437699680512, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1362, + "step": 1922 + }, + { + "epoch": 1.5359424920127795, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 1923 + }, + { + "epoch": 1.5367412140575079, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1347, + "step": 1924 + }, + { + "epoch": 1.5375399361022364, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1446, + "step": 1925 + }, + { + "epoch": 1.538338658146965, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1477, + "step": 1926 + }, + { + "epoch": 1.5391373801916934, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1345, + "step": 1927 + }, + { + "epoch": 1.5399361022364217, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1378, + "step": 1928 + }, + { + "epoch": 1.54073482428115, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1415, + "step": 1929 + }, + { + "epoch": 1.5415335463258786, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 1930 + }, + { + "epoch": 1.542332268370607, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1327, + "step": 1931 + }, + { + "epoch": 1.5431309904153354, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 1932 + }, + { + "epoch": 1.543929712460064, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 1933 + }, + { + "epoch": 1.5447284345047922, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 1934 + }, + { + "epoch": 1.5455271565495208, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1394, + "step": 1935 + }, + { + "epoch": 1.5463258785942493, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1366, + "step": 1936 + }, + { + "epoch": 1.5471246006389776, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1452, + "step": 1937 + }, + { + "epoch": 1.547923322683706, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1378, + "step": 1938 + }, + { + "epoch": 1.5487220447284344, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1366, + "step": 1939 + }, + { + "epoch": 1.549520766773163, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1398, + "step": 1940 + }, + { + "epoch": 1.5503194888178915, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.151, + "step": 1941 + }, + { + "epoch": 1.5511182108626198, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 1942 + }, + { + "epoch": 1.5519169329073481, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1384, + "step": 1943 + }, + { + "epoch": 1.5527156549520766, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1346, + "step": 1944 + }, + { + "epoch": 1.5535143769968052, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1448, + "step": 1945 + }, + { + "epoch": 1.5543130990415337, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 1946 + }, + { + "epoch": 1.555111821086262, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1457, + "step": 1947 + }, + { + "epoch": 1.5559105431309903, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1427, + "step": 1948 + }, + { + "epoch": 1.5567092651757188, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1349, + "step": 1949 + }, + { + "epoch": 1.5575079872204474, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1471, + "step": 1950 + }, + { + "epoch": 1.5583067092651757, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1459, + "step": 1951 + }, + { + "epoch": 1.5591054313099042, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1348, + "step": 1952 + }, + { + "epoch": 1.5599041533546325, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 1953 + }, + { + "epoch": 1.560702875399361, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1392, + "step": 1954 + }, + { + "epoch": 1.5615015974440896, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.142, + "step": 1955 + }, + { + "epoch": 1.5623003194888179, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1443, + "step": 1956 + }, + { + "epoch": 1.5630990415335462, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1394, + "step": 1957 + }, + { + "epoch": 1.5638977635782747, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 1958 + }, + { + "epoch": 1.5646964856230032, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.138, + "step": 1959 + }, + { + "epoch": 1.5654952076677318, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1411, + "step": 1960 + }, + { + "epoch": 1.56629392971246, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1376, + "step": 1961 + }, + { + "epoch": 1.5670926517571884, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1323, + "step": 1962 + }, + { + "epoch": 1.567891373801917, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 1963 + }, + { + "epoch": 1.5686900958466454, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.133, + "step": 1964 + }, + { + "epoch": 1.569488817891374, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 1965 + }, + { + "epoch": 1.5702875399361023, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1408, + "step": 1966 + }, + { + "epoch": 1.5710862619808306, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1433, + "step": 1967 + }, + { + "epoch": 1.571884984025559, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 1968 + }, + { + "epoch": 1.5726837060702876, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1418, + "step": 1969 + }, + { + "epoch": 1.573482428115016, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 1970 + }, + { + "epoch": 1.5742811501597445, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1532, + "step": 1971 + }, + { + "epoch": 1.5750798722044728, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1459, + "step": 1972 + }, + { + "epoch": 1.5758785942492013, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1457, + "step": 1973 + }, + { + "epoch": 1.5766773162939298, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1412, + "step": 1974 + }, + { + "epoch": 1.5774760383386581, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1405, + "step": 1975 + }, + { + "epoch": 1.5782747603833864, + "grad_norm": 0.39453125, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 1976 + }, + { + "epoch": 1.579073482428115, + "grad_norm": 0.4375, + "learning_rate": 0.0001, + "loss": 1.1363, + "step": 1977 + }, + { + "epoch": 1.5798722044728435, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 1978 + }, + { + "epoch": 1.580670926517572, + "grad_norm": 0.462890625, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 1979 + }, + { + "epoch": 1.5814696485623003, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.1367, + "step": 1980 + }, + { + "epoch": 1.5822683706070286, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1419, + "step": 1981 + }, + { + "epoch": 1.5830670926517572, + "grad_norm": 0.5546875, + "learning_rate": 0.0001, + "loss": 1.1393, + "step": 1982 + }, + { + "epoch": 1.5838658146964857, + "grad_norm": 0.58203125, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 1983 + }, + { + "epoch": 1.5846645367412142, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1386, + "step": 1984 + }, + { + "epoch": 1.5854632587859425, + "grad_norm": 0.388671875, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 1985 + }, + { + "epoch": 1.5862619808306708, + "grad_norm": 0.515625, + "learning_rate": 0.0001, + "loss": 1.1348, + "step": 1986 + }, + { + "epoch": 1.5870607028753994, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 1987 + }, + { + "epoch": 1.5878594249201279, + "grad_norm": 0.34765625, + "learning_rate": 0.0001, + "loss": 1.1495, + "step": 1988 + }, + { + "epoch": 1.5886581469648562, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1396, + "step": 1989 + }, + { + "epoch": 1.5894568690095847, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 1990 + }, + { + "epoch": 1.590255591054313, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1447, + "step": 1991 + }, + { + "epoch": 1.5910543130990416, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 1992 + }, + { + "epoch": 1.59185303514377, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1321, + "step": 1993 + }, + { + "epoch": 1.5926517571884984, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 1994 + }, + { + "epoch": 1.5934504792332267, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1479, + "step": 1995 + }, + { + "epoch": 1.5942492012779552, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 1996 + }, + { + "epoch": 1.5950479233226837, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 1997 + }, + { + "epoch": 1.5958466453674123, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1487, + "step": 1998 + }, + { + "epoch": 1.5966453674121406, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1435, + "step": 1999 + }, + { + "epoch": 1.5974440894568689, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1294, + "step": 2000 + }, + { + "epoch": 1.5982428115015974, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1386, + "step": 2001 + }, + { + "epoch": 1.599041533546326, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1357, + "step": 2002 + }, + { + "epoch": 1.5998402555910545, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1368, + "step": 2003 + }, + { + "epoch": 1.6006389776357828, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1377, + "step": 2004 + }, + { + "epoch": 1.601437699680511, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 2005 + }, + { + "epoch": 1.6022364217252396, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 2006 + }, + { + "epoch": 1.6030351437699681, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.147, + "step": 2007 + }, + { + "epoch": 1.6038338658146964, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1458, + "step": 2008 + }, + { + "epoch": 1.604632587859425, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.1417, + "step": 2009 + }, + { + "epoch": 1.6054313099041533, + "grad_norm": 0.66796875, + "learning_rate": 0.0001, + "loss": 1.1383, + "step": 2010 + }, + { + "epoch": 1.6062300319488818, + "grad_norm": 0.5703125, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 2011 + }, + { + "epoch": 1.6070287539936103, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 2012 + }, + { + "epoch": 1.6078274760383386, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 2013 + }, + { + "epoch": 1.608626198083067, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1327, + "step": 2014 + }, + { + "epoch": 1.6094249201277955, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 2015 + }, + { + "epoch": 1.610223642172524, + "grad_norm": 0.47265625, + "learning_rate": 0.0001, + "loss": 1.1411, + "step": 2016 + }, + { + "epoch": 1.6110223642172525, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 2017 + }, + { + "epoch": 1.6118210862619808, + "grad_norm": 0.59375, + "learning_rate": 0.0001, + "loss": 1.1454, + "step": 2018 + }, + { + "epoch": 1.6126198083067091, + "grad_norm": 0.63671875, + "learning_rate": 0.0001, + "loss": 1.1317, + "step": 2019 + }, + { + "epoch": 1.6134185303514377, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 2020 + }, + { + "epoch": 1.6142172523961662, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.1514, + "step": 2021 + }, + { + "epoch": 1.6150159744408947, + "grad_norm": 0.44921875, + "learning_rate": 0.0001, + "loss": 1.1418, + "step": 2022 + }, + { + "epoch": 1.615814696485623, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1431, + "step": 2023 + }, + { + "epoch": 1.6166134185303513, + "grad_norm": 0.5625, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 2024 + }, + { + "epoch": 1.6174121405750799, + "grad_norm": 0.357421875, + "learning_rate": 0.0001, + "loss": 1.1368, + "step": 2025 + }, + { + "epoch": 1.6182108626198084, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1363, + "step": 2026 + }, + { + "epoch": 1.6190095846645367, + "grad_norm": 0.431640625, + "learning_rate": 0.0001, + "loss": 1.1466, + "step": 2027 + }, + { + "epoch": 1.619808306709265, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 2028 + }, + { + "epoch": 1.6206070287539935, + "grad_norm": 0.380859375, + "learning_rate": 0.0001, + "loss": 1.1409, + "step": 2029 + }, + { + "epoch": 1.621405750798722, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1399, + "step": 2030 + }, + { + "epoch": 1.6222044728434506, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 2031 + }, + { + "epoch": 1.623003194888179, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1381, + "step": 2032 + }, + { + "epoch": 1.6238019169329072, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 2033 + }, + { + "epoch": 1.6246006389776357, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1323, + "step": 2034 + }, + { + "epoch": 1.6253993610223643, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1438, + "step": 2035 + }, + { + "epoch": 1.6261980830670928, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1443, + "step": 2036 + }, + { + "epoch": 1.626996805111821, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.134, + "step": 2037 + }, + { + "epoch": 1.6277955271565494, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1468, + "step": 2038 + }, + { + "epoch": 1.628594249201278, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1363, + "step": 2039 + }, + { + "epoch": 1.6293929712460065, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1451, + "step": 2040 + }, + { + "epoch": 1.630191693290735, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 2041 + }, + { + "epoch": 1.6309904153354633, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1362, + "step": 2042 + }, + { + "epoch": 1.6317891373801916, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 2043 + }, + { + "epoch": 1.6325878594249201, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.137, + "step": 2044 + }, + { + "epoch": 1.6333865814696487, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1361, + "step": 2045 + }, + { + "epoch": 1.634185303514377, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 2046 + }, + { + "epoch": 1.6349840255591053, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 2047 + }, + { + "epoch": 1.6357827476038338, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1381, + "step": 2048 + }, + { + "epoch": 1.6365814696485623, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1452, + "step": 2049 + }, + { + "epoch": 1.6373801916932909, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1347, + "step": 2050 + }, + { + "epoch": 1.6381789137380192, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 2051 + }, + { + "epoch": 1.6389776357827475, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 2052 + }, + { + "epoch": 1.639776357827476, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 2053 + }, + { + "epoch": 1.6405750798722045, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 2054 + }, + { + "epoch": 1.641373801916933, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1345, + "step": 2055 + }, + { + "epoch": 1.6421725239616614, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1406, + "step": 2056 + }, + { + "epoch": 1.6429712460063897, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 2057 + }, + { + "epoch": 1.6437699680511182, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1418, + "step": 2058 + }, + { + "epoch": 1.6445686900958467, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1409, + "step": 2059 + }, + { + "epoch": 1.645367412140575, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 2060 + }, + { + "epoch": 1.6461661341853036, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 2061 + }, + { + "epoch": 1.6469648562300319, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1373, + "step": 2062 + }, + { + "epoch": 1.6477635782747604, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1385, + "step": 2063 + }, + { + "epoch": 1.648562300319489, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 2064 + }, + { + "epoch": 1.6493610223642172, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1384, + "step": 2065 + }, + { + "epoch": 1.6501597444089455, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1374, + "step": 2066 + }, + { + "epoch": 1.650958466453674, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1359, + "step": 2067 + }, + { + "epoch": 1.6517571884984026, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1325, + "step": 2068 + }, + { + "epoch": 1.6525559105431311, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1397, + "step": 2069 + }, + { + "epoch": 1.6533546325878594, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2070 + }, + { + "epoch": 1.6541533546325877, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 2071 + }, + { + "epoch": 1.6549520766773163, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1381, + "step": 2072 + }, + { + "epoch": 1.6557507987220448, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.144, + "step": 2073 + }, + { + "epoch": 1.6565495207667733, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1373, + "step": 2074 + }, + { + "epoch": 1.6573482428115016, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1301, + "step": 2075 + }, + { + "epoch": 1.65814696485623, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 2076 + }, + { + "epoch": 1.6589456869009584, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 2077 + }, + { + "epoch": 1.659744408945687, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2078 + }, + { + "epoch": 1.6605431309904153, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 2079 + }, + { + "epoch": 1.6613418530351438, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2080 + }, + { + "epoch": 1.6621405750798721, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 2081 + }, + { + "epoch": 1.6629392971246006, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1395, + "step": 2082 + }, + { + "epoch": 1.6637380191693292, + "grad_norm": 0.470703125, + "learning_rate": 0.0001, + "loss": 1.1413, + "step": 2083 + }, + { + "epoch": 1.6645367412140575, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1392, + "step": 2084 + }, + { + "epoch": 1.6653354632587858, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 2085 + }, + { + "epoch": 1.6661341853035143, + "grad_norm": 0.462890625, + "learning_rate": 0.0001, + "loss": 1.1329, + "step": 2086 + }, + { + "epoch": 1.6669329073482428, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1319, + "step": 2087 + }, + { + "epoch": 1.6677316293929714, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1329, + "step": 2088 + }, + { + "epoch": 1.6685303514376997, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 2089 + }, + { + "epoch": 1.669329073482428, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 2090 + }, + { + "epoch": 1.6701277955271565, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 2091 + }, + { + "epoch": 1.670926517571885, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1434, + "step": 2092 + }, + { + "epoch": 1.6717252396166136, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 2093 + }, + { + "epoch": 1.6725239616613419, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1422, + "step": 2094 + }, + { + "epoch": 1.6733226837060702, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 2095 + }, + { + "epoch": 1.6741214057507987, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 2096 + }, + { + "epoch": 1.6749201277955272, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1428, + "step": 2097 + }, + { + "epoch": 1.6757188498402555, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1376, + "step": 2098 + }, + { + "epoch": 1.676517571884984, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1407, + "step": 2099 + }, + { + "epoch": 1.6773162939297124, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 2100 + }, + { + "epoch": 1.678115015974441, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 2101 + }, + { + "epoch": 1.6789137380191694, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 2102 + }, + { + "epoch": 1.6797124600638977, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1432, + "step": 2103 + }, + { + "epoch": 1.680511182108626, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 2104 + }, + { + "epoch": 1.6813099041533546, + "grad_norm": 0.390625, + "learning_rate": 0.0001, + "loss": 1.138, + "step": 2105 + }, + { + "epoch": 1.682108626198083, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1418, + "step": 2106 + }, + { + "epoch": 1.6829073482428116, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1437, + "step": 2107 + }, + { + "epoch": 1.68370607028754, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 2108 + }, + { + "epoch": 1.6845047923322682, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1419, + "step": 2109 + }, + { + "epoch": 1.6853035143769968, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 2110 + }, + { + "epoch": 1.6861022364217253, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 2111 + }, + { + "epoch": 1.6869009584664538, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 2112 + }, + { + "epoch": 1.6876996805111821, + "grad_norm": 0.353515625, + "learning_rate": 0.0001, + "loss": 1.1408, + "step": 2113 + }, + { + "epoch": 1.6884984025559104, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1373, + "step": 2114 + }, + { + "epoch": 1.689297124600639, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1361, + "step": 2115 + }, + { + "epoch": 1.6900958466453675, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1413, + "step": 2116 + }, + { + "epoch": 1.6908945686900958, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 2117 + }, + { + "epoch": 1.6916932907348243, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 2118 + }, + { + "epoch": 1.6924920127795526, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 2119 + }, + { + "epoch": 1.6932907348242812, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1471, + "step": 2120 + }, + { + "epoch": 1.6940894568690097, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1422, + "step": 2121 + }, + { + "epoch": 1.694888178913738, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1387, + "step": 2122 + }, + { + "epoch": 1.6956869009584663, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1332, + "step": 2123 + }, + { + "epoch": 1.6964856230031948, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 2124 + }, + { + "epoch": 1.6972843450479234, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1414, + "step": 2125 + }, + { + "epoch": 1.6980830670926519, + "grad_norm": 0.361328125, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 2126 + }, + { + "epoch": 1.6988817891373802, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 2127 + }, + { + "epoch": 1.6996805111821085, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1466, + "step": 2128 + }, + { + "epoch": 1.700479233226837, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1419, + "step": 2129 + }, + { + "epoch": 1.7012779552715656, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 2130 + }, + { + "epoch": 1.702076677316294, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1381, + "step": 2131 + }, + { + "epoch": 1.7028753993610224, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 2132 + }, + { + "epoch": 1.7036741214057507, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 2133 + }, + { + "epoch": 1.7044728434504792, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1333, + "step": 2134 + }, + { + "epoch": 1.7052715654952078, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1332, + "step": 2135 + }, + { + "epoch": 1.706070287539936, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 2136 + }, + { + "epoch": 1.7068690095846646, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1393, + "step": 2137 + }, + { + "epoch": 1.707667731629393, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 2138 + }, + { + "epoch": 1.7084664536741214, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 2139 + }, + { + "epoch": 1.70926517571885, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1451, + "step": 2140 + }, + { + "epoch": 1.7100638977635783, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1337, + "step": 2141 + }, + { + "epoch": 1.7108626198083066, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 2142 + }, + { + "epoch": 1.711661341853035, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1457, + "step": 2143 + }, + { + "epoch": 1.7124600638977636, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1375, + "step": 2144 + }, + { + "epoch": 1.7132587859424921, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1407, + "step": 2145 + }, + { + "epoch": 1.7140575079872205, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2146 + }, + { + "epoch": 1.7148562300319488, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1415, + "step": 2147 + }, + { + "epoch": 1.7156549520766773, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1365, + "step": 2148 + }, + { + "epoch": 1.7164536741214058, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 2149 + }, + { + "epoch": 1.7172523961661343, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 2150 + }, + { + "epoch": 1.7180511182108626, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 2151 + }, + { + "epoch": 1.718849840255591, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1294, + "step": 2152 + }, + { + "epoch": 1.7196485623003195, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1381, + "step": 2153 + }, + { + "epoch": 1.720447284345048, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1305, + "step": 2154 + }, + { + "epoch": 1.7212460063897763, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 2155 + }, + { + "epoch": 1.7220447284345048, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 2156 + }, + { + "epoch": 1.7228434504792332, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 2157 + }, + { + "epoch": 1.7236421725239617, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1375, + "step": 2158 + }, + { + "epoch": 1.7244408945686902, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 2159 + }, + { + "epoch": 1.7252396166134185, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1367, + "step": 2160 + }, + { + "epoch": 1.7260383386581468, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1443, + "step": 2161 + }, + { + "epoch": 1.7268370607028753, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 2162 + }, + { + "epoch": 1.7276357827476039, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 2163 + }, + { + "epoch": 1.7284345047923324, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 2164 + }, + { + "epoch": 1.7292332268370607, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1375, + "step": 2165 + }, + { + "epoch": 1.730031948881789, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1327, + "step": 2166 + }, + { + "epoch": 1.7308306709265175, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 2167 + }, + { + "epoch": 1.731629392971246, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1327, + "step": 2168 + }, + { + "epoch": 1.7324281150159746, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 2169 + }, + { + "epoch": 1.733226837060703, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 2170 + }, + { + "epoch": 1.7340255591054312, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2171 + }, + { + "epoch": 1.7348242811501597, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 2172 + }, + { + "epoch": 1.7356230031948883, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1346, + "step": 2173 + }, + { + "epoch": 1.7364217252396166, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 2174 + }, + { + "epoch": 1.7372204472843449, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 2175 + }, + { + "epoch": 1.7380191693290734, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1391, + "step": 2176 + }, + { + "epoch": 1.738817891373802, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 2177 + }, + { + "epoch": 1.7396166134185305, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 2178 + }, + { + "epoch": 1.7404153354632588, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1417, + "step": 2179 + }, + { + "epoch": 1.741214057507987, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1332, + "step": 2180 + }, + { + "epoch": 1.7420127795527156, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 2181 + }, + { + "epoch": 1.7428115015974441, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1369, + "step": 2182 + }, + { + "epoch": 1.7436102236421727, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1429, + "step": 2183 + }, + { + "epoch": 1.744408945686901, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2184 + }, + { + "epoch": 1.7452076677316293, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 2185 + }, + { + "epoch": 1.7460063897763578, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 2186 + }, + { + "epoch": 1.7468051118210863, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 2187 + }, + { + "epoch": 1.7476038338658149, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 2188 + }, + { + "epoch": 1.7484025559105432, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1333, + "step": 2189 + }, + { + "epoch": 1.7492012779552715, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1321, + "step": 2190 + }, + { + "epoch": 1.75, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1394, + "step": 2191 + }, + { + "epoch": 1.7507987220447285, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1362, + "step": 2192 + }, + { + "epoch": 1.7515974440894568, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1405, + "step": 2193 + }, + { + "epoch": 1.7523961661341851, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 2194 + }, + { + "epoch": 1.7531948881789137, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 2195 + }, + { + "epoch": 1.7539936102236422, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 2196 + }, + { + "epoch": 1.7547923322683707, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1361, + "step": 2197 + }, + { + "epoch": 1.755591054313099, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 2198 + }, + { + "epoch": 1.7563897763578273, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1323, + "step": 2199 + }, + { + "epoch": 1.7571884984025559, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1334, + "step": 2200 + }, + { + "epoch": 1.7579872204472844, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 2201 + }, + { + "epoch": 1.758785942492013, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1397, + "step": 2202 + }, + { + "epoch": 1.7595846645367412, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 2203 + }, + { + "epoch": 1.7603833865814695, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 2204 + }, + { + "epoch": 1.761182108626198, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1414, + "step": 2205 + }, + { + "epoch": 1.7619808306709266, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1373, + "step": 2206 + }, + { + "epoch": 1.7627795527156551, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.137, + "step": 2207 + }, + { + "epoch": 1.7635782747603834, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.135, + "step": 2208 + }, + { + "epoch": 1.7643769968051117, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1346, + "step": 2209 + }, + { + "epoch": 1.7651757188498403, + "grad_norm": 0.392578125, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 2210 + }, + { + "epoch": 1.7659744408945688, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 2211 + }, + { + "epoch": 1.766773162939297, + "grad_norm": 0.37890625, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 2212 + }, + { + "epoch": 1.7675718849840254, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1332, + "step": 2213 + }, + { + "epoch": 1.768370607028754, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1333, + "step": 2214 + }, + { + "epoch": 1.7691693290734825, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 2215 + }, + { + "epoch": 1.769968051118211, + "grad_norm": 0.34765625, + "learning_rate": 0.0001, + "loss": 1.1357, + "step": 2216 + }, + { + "epoch": 1.7707667731629393, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 2217 + }, + { + "epoch": 1.7715654952076676, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1396, + "step": 2218 + }, + { + "epoch": 1.7723642172523961, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 2219 + }, + { + "epoch": 1.7731629392971247, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1295, + "step": 2220 + }, + { + "epoch": 1.7739616613418532, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 2221 + }, + { + "epoch": 1.7747603833865815, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1311, + "step": 2222 + }, + { + "epoch": 1.7755591054313098, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2223 + }, + { + "epoch": 1.7763578274760383, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1387, + "step": 2224 + }, + { + "epoch": 1.7771565495207668, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 2225 + }, + { + "epoch": 1.7779552715654952, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 2226 + }, + { + "epoch": 1.7787539936102237, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2227 + }, + { + "epoch": 1.779552715654952, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1501, + "step": 2228 + }, + { + "epoch": 1.7803514376996805, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 2229 + }, + { + "epoch": 1.781150159744409, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1398, + "step": 2230 + }, + { + "epoch": 1.7819488817891374, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 2231 + }, + { + "epoch": 1.7827476038338657, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 2232 + }, + { + "epoch": 1.7835463258785942, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1382, + "step": 2233 + }, + { + "epoch": 1.7843450479233227, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1422, + "step": 2234 + }, + { + "epoch": 1.7851437699680512, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1323, + "step": 2235 + }, + { + "epoch": 1.7859424920127795, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1395, + "step": 2236 + }, + { + "epoch": 1.7867412140575079, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 2237 + }, + { + "epoch": 1.7875399361022364, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2238 + }, + { + "epoch": 1.788338658146965, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1393, + "step": 2239 + }, + { + "epoch": 1.7891373801916934, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 2240 + }, + { + "epoch": 1.7899361022364217, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.15, + "step": 2241 + }, + { + "epoch": 1.79073482428115, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.137, + "step": 2242 + }, + { + "epoch": 1.7915335463258786, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1329, + "step": 2243 + }, + { + "epoch": 1.792332268370607, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.135, + "step": 2244 + }, + { + "epoch": 1.7931309904153354, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1362, + "step": 2245 + }, + { + "epoch": 1.793929712460064, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1352, + "step": 2246 + }, + { + "epoch": 1.7947284345047922, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1435, + "step": 2247 + }, + { + "epoch": 1.7955271565495208, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 2248 + }, + { + "epoch": 1.7963258785942493, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 2249 + }, + { + "epoch": 1.7971246006389776, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1413, + "step": 2250 + }, + { + "epoch": 1.797923322683706, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1335, + "step": 2251 + }, + { + "epoch": 1.7987220447284344, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 2252 + }, + { + "epoch": 1.799520766773163, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 2253 + }, + { + "epoch": 1.8003194888178915, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1412, + "step": 2254 + }, + { + "epoch": 1.8011182108626198, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1295, + "step": 2255 + }, + { + "epoch": 1.8019169329073481, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1337, + "step": 2256 + }, + { + "epoch": 1.8027156549520766, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1382, + "step": 2257 + }, + { + "epoch": 1.8035143769968052, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1423, + "step": 2258 + }, + { + "epoch": 1.8043130990415337, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.144, + "step": 2259 + }, + { + "epoch": 1.805111821086262, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 2260 + }, + { + "epoch": 1.8059105431309903, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 2261 + }, + { + "epoch": 1.8067092651757188, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 2262 + }, + { + "epoch": 1.8075079872204474, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.138, + "step": 2263 + }, + { + "epoch": 1.8083067092651757, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1323, + "step": 2264 + }, + { + "epoch": 1.8091054313099042, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1327, + "step": 2265 + }, + { + "epoch": 1.8099041533546325, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 2266 + }, + { + "epoch": 1.810702875399361, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 2267 + }, + { + "epoch": 1.8115015974440896, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 2268 + }, + { + "epoch": 1.8123003194888179, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 2269 + }, + { + "epoch": 1.8130990415335462, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 2270 + }, + { + "epoch": 1.8138977635782747, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 2271 + }, + { + "epoch": 1.8146964856230032, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1367, + "step": 2272 + }, + { + "epoch": 1.8154952076677318, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 2273 + }, + { + "epoch": 1.81629392971246, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.137, + "step": 2274 + }, + { + "epoch": 1.8170926517571884, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 2275 + }, + { + "epoch": 1.817891373801917, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.133, + "step": 2276 + }, + { + "epoch": 1.8186900958466454, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 2277 + }, + { + "epoch": 1.819488817891374, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2278 + }, + { + "epoch": 1.8202875399361023, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1302, + "step": 2279 + }, + { + "epoch": 1.8210862619808306, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1407, + "step": 2280 + }, + { + "epoch": 1.821884984025559, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 2281 + }, + { + "epoch": 1.8226837060702876, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 2282 + }, + { + "epoch": 1.823482428115016, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1321, + "step": 2283 + }, + { + "epoch": 1.8242811501597445, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.142, + "step": 2284 + }, + { + "epoch": 1.8250798722044728, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2285 + }, + { + "epoch": 1.8258785942492013, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1347, + "step": 2286 + }, + { + "epoch": 1.8266773162939298, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1397, + "step": 2287 + }, + { + "epoch": 1.8274760383386581, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 2288 + }, + { + "epoch": 1.8282747603833864, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1385, + "step": 2289 + }, + { + "epoch": 1.829073482428115, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 2290 + }, + { + "epoch": 1.8298722044728435, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1413, + "step": 2291 + }, + { + "epoch": 1.830670926517572, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1325, + "step": 2292 + }, + { + "epoch": 1.8314696485623003, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 2293 + }, + { + "epoch": 1.8322683706070286, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1335, + "step": 2294 + }, + { + "epoch": 1.8330670926517572, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 2295 + }, + { + "epoch": 1.8338658146964857, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1368, + "step": 2296 + }, + { + "epoch": 1.8346645367412142, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1348, + "step": 2297 + }, + { + "epoch": 1.8354632587859425, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 2298 + }, + { + "epoch": 1.8362619808306708, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.138, + "step": 2299 + }, + { + "epoch": 1.8370607028753994, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1404, + "step": 2300 + }, + { + "epoch": 1.8378594249201279, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 2301 + }, + { + "epoch": 1.8386581469648562, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1347, + "step": 2302 + }, + { + "epoch": 1.8394568690095847, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.137, + "step": 2303 + }, + { + "epoch": 1.840255591054313, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1432, + "step": 2304 + }, + { + "epoch": 1.8410543130990416, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1294, + "step": 2305 + }, + { + "epoch": 1.84185303514377, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 2306 + }, + { + "epoch": 1.8426517571884984, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1352, + "step": 2307 + }, + { + "epoch": 1.8434504792332267, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1349, + "step": 2308 + }, + { + "epoch": 1.8442492012779552, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1302, + "step": 2309 + }, + { + "epoch": 1.8450479233226837, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 2310 + }, + { + "epoch": 1.8458466453674123, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 2311 + }, + { + "epoch": 1.8466453674121406, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 2312 + }, + { + "epoch": 1.8474440894568689, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.137, + "step": 2313 + }, + { + "epoch": 1.8482428115015974, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1365, + "step": 2314 + }, + { + "epoch": 1.849041533546326, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 2315 + }, + { + "epoch": 1.8498402555910545, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1334, + "step": 2316 + }, + { + "epoch": 1.8506389776357828, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1417, + "step": 2317 + }, + { + "epoch": 1.851437699680511, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1393, + "step": 2318 + }, + { + "epoch": 1.8522364217252396, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 2319 + }, + { + "epoch": 1.8530351437699681, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1287, + "step": 2320 + }, + { + "epoch": 1.8538338658146964, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 2321 + }, + { + "epoch": 1.854632587859425, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1476, + "step": 2322 + }, + { + "epoch": 1.8554313099041533, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1294, + "step": 2323 + }, + { + "epoch": 1.8562300319488818, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1287, + "step": 2324 + }, + { + "epoch": 1.8570287539936103, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 2325 + }, + { + "epoch": 1.8578274760383386, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.134, + "step": 2326 + }, + { + "epoch": 1.858626198083067, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 2327 + }, + { + "epoch": 1.8594249201277955, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 2328 + }, + { + "epoch": 1.860223642172524, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1327, + "step": 2329 + }, + { + "epoch": 1.8610223642172525, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 2330 + }, + { + "epoch": 1.8618210862619808, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 2331 + }, + { + "epoch": 1.8626198083067091, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2332 + }, + { + "epoch": 1.8634185303514377, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 2333 + }, + { + "epoch": 1.8642172523961662, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1369, + "step": 2334 + }, + { + "epoch": 1.8650159744408947, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1355, + "step": 2335 + }, + { + "epoch": 1.865814696485623, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 2336 + }, + { + "epoch": 1.8666134185303513, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1323, + "step": 2337 + }, + { + "epoch": 1.8674121405750799, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 2338 + }, + { + "epoch": 1.8682108626198084, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 2339 + }, + { + "epoch": 1.8690095846645367, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1377, + "step": 2340 + }, + { + "epoch": 1.869808306709265, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1292, + "step": 2341 + }, + { + "epoch": 1.8706070287539935, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.128, + "step": 2342 + }, + { + "epoch": 1.871405750798722, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1283, + "step": 2343 + }, + { + "epoch": 1.8722044728434506, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1369, + "step": 2344 + }, + { + "epoch": 1.873003194888179, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 2345 + }, + { + "epoch": 1.8738019169329072, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1425, + "step": 2346 + }, + { + "epoch": 1.8746006389776357, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2347 + }, + { + "epoch": 1.8753993610223643, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1352, + "step": 2348 + }, + { + "epoch": 1.8761980830670928, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 2349 + }, + { + "epoch": 1.876996805111821, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 2350 + }, + { + "epoch": 1.8777955271565494, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 2351 + }, + { + "epoch": 1.878594249201278, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 2352 + }, + { + "epoch": 1.8793929712460065, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 2353 + }, + { + "epoch": 1.880191693290735, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 2354 + }, + { + "epoch": 1.8809904153354633, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 2355 + }, + { + "epoch": 1.8817891373801916, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1387, + "step": 2356 + }, + { + "epoch": 1.8825878594249201, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1425, + "step": 2357 + }, + { + "epoch": 1.8833865814696487, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 2358 + }, + { + "epoch": 1.884185303514377, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1444, + "step": 2359 + }, + { + "epoch": 1.8849840255591053, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1327, + "step": 2360 + }, + { + "epoch": 1.8857827476038338, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 2361 + }, + { + "epoch": 1.8865814696485623, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 2362 + }, + { + "epoch": 1.8873801916932909, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1345, + "step": 2363 + }, + { + "epoch": 1.8881789137380192, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1386, + "step": 2364 + }, + { + "epoch": 1.8889776357827475, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1357, + "step": 2365 + }, + { + "epoch": 1.889776357827476, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 2366 + }, + { + "epoch": 1.8905750798722045, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 2367 + }, + { + "epoch": 1.891373801916933, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 2368 + }, + { + "epoch": 1.8921725239616614, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 2369 + }, + { + "epoch": 1.8929712460063897, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1447, + "step": 2370 + }, + { + "epoch": 1.8937699680511182, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 2371 + }, + { + "epoch": 1.8945686900958467, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 2372 + }, + { + "epoch": 1.895367412140575, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1287, + "step": 2373 + }, + { + "epoch": 1.8961661341853036, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 2374 + }, + { + "epoch": 1.8969648562300319, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1412, + "step": 2375 + }, + { + "epoch": 1.8977635782747604, + "grad_norm": 0.40234375, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 2376 + }, + { + "epoch": 1.898562300319489, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 2377 + }, + { + "epoch": 1.8993610223642172, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 2378 + }, + { + "epoch": 1.9001597444089455, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1269, + "step": 2379 + }, + { + "epoch": 1.900958466453674, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1357, + "step": 2380 + }, + { + "epoch": 1.9017571884984026, + "grad_norm": 0.337890625, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 2381 + }, + { + "epoch": 1.9025559105431311, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1366, + "step": 2382 + }, + { + "epoch": 1.9033546325878594, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1363, + "step": 2383 + }, + { + "epoch": 1.9041533546325877, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 2384 + }, + { + "epoch": 1.9049520766773163, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 2385 + }, + { + "epoch": 1.9057507987220448, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1369, + "step": 2386 + }, + { + "epoch": 1.9065495207667733, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 2387 + }, + { + "epoch": 1.9073482428115016, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 2388 + }, + { + "epoch": 1.90814696485623, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 2389 + }, + { + "epoch": 1.9089456869009584, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1361, + "step": 2390 + }, + { + "epoch": 1.909744408945687, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1333, + "step": 2391 + }, + { + "epoch": 1.9105431309904153, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1367, + "step": 2392 + }, + { + "epoch": 1.9113418530351438, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 2393 + }, + { + "epoch": 1.9121405750798721, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1416, + "step": 2394 + }, + { + "epoch": 1.9129392971246006, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 2395 + }, + { + "epoch": 1.9137380191693292, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 2396 + }, + { + "epoch": 1.9145367412140575, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 2397 + }, + { + "epoch": 1.9153354632587858, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1337, + "step": 2398 + }, + { + "epoch": 1.9161341853035143, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1374, + "step": 2399 + }, + { + "epoch": 1.9169329073482428, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2400 + }, + { + "epoch": 1.9177316293929714, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 2401 + }, + { + "epoch": 1.9185303514376997, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 2402 + }, + { + "epoch": 1.919329073482428, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 2403 + }, + { + "epoch": 1.9201277955271565, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 2404 + }, + { + "epoch": 1.920926517571885, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2405 + }, + { + "epoch": 1.9217252396166136, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1319, + "step": 2406 + }, + { + "epoch": 1.9225239616613419, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1378, + "step": 2407 + }, + { + "epoch": 1.9233226837060702, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1385, + "step": 2408 + }, + { + "epoch": 1.9241214057507987, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 2409 + }, + { + "epoch": 1.9249201277955272, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 2410 + }, + { + "epoch": 1.9257188498402555, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1329, + "step": 2411 + }, + { + "epoch": 1.926517571884984, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1454, + "step": 2412 + }, + { + "epoch": 1.9273162939297124, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1428, + "step": 2413 + }, + { + "epoch": 1.928115015974441, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1385, + "step": 2414 + }, + { + "epoch": 1.9289137380191694, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 2415 + }, + { + "epoch": 1.9297124600638977, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1295, + "step": 2416 + }, + { + "epoch": 1.930511182108626, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 2417 + }, + { + "epoch": 1.9313099041533546, + "grad_norm": 0.345703125, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 2418 + }, + { + "epoch": 1.932108626198083, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 2419 + }, + { + "epoch": 1.9329073482428116, + "grad_norm": 0.5078125, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 2420 + }, + { + "epoch": 1.93370607028754, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1319, + "step": 2421 + }, + { + "epoch": 1.9345047923322682, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.1408, + "step": 2422 + }, + { + "epoch": 1.9353035143769968, + "grad_norm": 0.703125, + "learning_rate": 0.0001, + "loss": 1.1367, + "step": 2423 + }, + { + "epoch": 1.9361022364217253, + "grad_norm": 0.9609375, + "learning_rate": 0.0001, + "loss": 1.1397, + "step": 2424 + }, + { + "epoch": 1.9369009584664538, + "grad_norm": 1.1953125, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 2425 + }, + { + "epoch": 1.9376996805111821, + "grad_norm": 0.478515625, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2426 + }, + { + "epoch": 1.9384984025559104, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1386, + "step": 2427 + }, + { + "epoch": 1.939297124600639, + "grad_norm": 0.69921875, + "learning_rate": 0.0001, + "loss": 1.1369, + "step": 2428 + }, + { + "epoch": 1.9400958466453675, + "grad_norm": 0.53515625, + "learning_rate": 0.0001, + "loss": 1.1383, + "step": 2429 + }, + { + "epoch": 1.9408945686900958, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 2430 + }, + { + "epoch": 1.9416932907348243, + "grad_norm": 1.046875, + "learning_rate": 0.0001, + "loss": 1.1463, + "step": 2431 + }, + { + "epoch": 1.9424920127795526, + "grad_norm": 1.3203125, + "learning_rate": 0.0001, + "loss": 1.1457, + "step": 2432 + }, + { + "epoch": 1.9432907348242812, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1382, + "step": 2433 + }, + { + "epoch": 1.9440894568690097, + "grad_norm": 0.6484375, + "learning_rate": 0.0001, + "loss": 1.1366, + "step": 2434 + }, + { + "epoch": 1.944888178913738, + "grad_norm": 0.734375, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 2435 + }, + { + "epoch": 1.9456869009584663, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 2436 + }, + { + "epoch": 1.9464856230031948, + "grad_norm": 0.5703125, + "learning_rate": 0.0001, + "loss": 1.1363, + "step": 2437 + }, + { + "epoch": 1.9472843450479234, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 2438 + }, + { + "epoch": 1.9480830670926519, + "grad_norm": 0.4453125, + "learning_rate": 0.0001, + "loss": 1.1398, + "step": 2439 + }, + { + "epoch": 1.9488817891373802, + "grad_norm": 0.8125, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 2440 + }, + { + "epoch": 1.9496805111821085, + "grad_norm": 0.40625, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 2441 + }, + { + "epoch": 1.950479233226837, + "grad_norm": 0.318359375, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 2442 + }, + { + "epoch": 1.9512779552715656, + "grad_norm": 0.5078125, + "learning_rate": 0.0001, + "loss": 1.1412, + "step": 2443 + }, + { + "epoch": 1.952076677316294, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 2444 + }, + { + "epoch": 1.9528753993610224, + "grad_norm": 0.5859375, + "learning_rate": 0.0001, + "loss": 1.1374, + "step": 2445 + }, + { + "epoch": 1.9536741214057507, + "grad_norm": 0.51171875, + "learning_rate": 0.0001, + "loss": 1.1382, + "step": 2446 + }, + { + "epoch": 1.9544728434504792, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 2447 + }, + { + "epoch": 1.9552715654952078, + "grad_norm": 0.69921875, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2448 + }, + { + "epoch": 1.956070287539936, + "grad_norm": 0.318359375, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 2449 + }, + { + "epoch": 1.9568690095846646, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1406, + "step": 2450 + }, + { + "epoch": 1.957667731629393, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.128, + "step": 2451 + }, + { + "epoch": 1.9584664536741214, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 2452 + }, + { + "epoch": 1.95926517571885, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1337, + "step": 2453 + }, + { + "epoch": 1.9600638977635783, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 2454 + }, + { + "epoch": 1.9608626198083066, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1356, + "step": 2455 + }, + { + "epoch": 1.961661341853035, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 2456 + }, + { + "epoch": 1.9624600638977636, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 2457 + }, + { + "epoch": 1.9632587859424921, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1359, + "step": 2458 + }, + { + "epoch": 1.9640575079872205, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1363, + "step": 2459 + }, + { + "epoch": 1.9648562300319488, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 2460 + }, + { + "epoch": 1.9656549520766773, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 2461 + }, + { + "epoch": 1.9664536741214058, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 2462 + }, + { + "epoch": 1.9672523961661343, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 2463 + }, + { + "epoch": 1.9680511182108626, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2464 + }, + { + "epoch": 1.968849840255591, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 2465 + }, + { + "epoch": 1.9696485623003195, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 2466 + }, + { + "epoch": 1.970447284345048, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 2467 + }, + { + "epoch": 1.9712460063897763, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 2468 + }, + { + "epoch": 1.9720447284345048, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1396, + "step": 2469 + }, + { + "epoch": 1.9728434504792332, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1328, + "step": 2470 + }, + { + "epoch": 1.9736421725239617, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 2471 + }, + { + "epoch": 1.9744408945686902, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 2472 + }, + { + "epoch": 1.9752396166134185, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 2473 + }, + { + "epoch": 1.9760383386581468, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1407, + "step": 2474 + }, + { + "epoch": 1.9768370607028753, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 2475 + }, + { + "epoch": 1.9776357827476039, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 2476 + }, + { + "epoch": 1.9784345047923324, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 2477 + }, + { + "epoch": 1.9792332268370607, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 2478 + }, + { + "epoch": 1.980031948881789, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 2479 + }, + { + "epoch": 1.9808306709265175, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 2480 + }, + { + "epoch": 1.981629392971246, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1405, + "step": 2481 + }, + { + "epoch": 1.9824281150159746, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 2482 + }, + { + "epoch": 1.983226837060703, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1391, + "step": 2483 + }, + { + "epoch": 1.9840255591054312, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1426, + "step": 2484 + }, + { + "epoch": 1.9848242811501597, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 2485 + }, + { + "epoch": 1.9856230031948883, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1442, + "step": 2486 + }, + { + "epoch": 1.9864217252396166, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1377, + "step": 2487 + }, + { + "epoch": 1.9872204472843449, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1327, + "step": 2488 + }, + { + "epoch": 1.9880191693290734, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 2489 + }, + { + "epoch": 1.988817891373802, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1387, + "step": 2490 + }, + { + "epoch": 1.9896166134185305, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1325, + "step": 2491 + }, + { + "epoch": 1.9904153354632588, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 2492 + }, + { + "epoch": 1.991214057507987, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 2493 + }, + { + "epoch": 1.9920127795527156, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 2494 + }, + { + "epoch": 1.9928115015974441, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 2495 + }, + { + "epoch": 1.9936102236421727, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.135, + "step": 2496 + }, + { + "epoch": 1.994408945686901, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 2497 + }, + { + "epoch": 1.9952076677316293, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 2498 + }, + { + "epoch": 1.9960063897763578, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1294, + "step": 2499 + }, + { + "epoch": 1.9968051118210863, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 2500 + }, + { + "epoch": 1.9976038338658149, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1408, + "step": 2501 + }, + { + "epoch": 1.9984025559105432, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.143, + "step": 2502 + }, + { + "epoch": 1.9992012779552715, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1366, + "step": 2503 + }, + { + "epoch": 2.0, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 2504 + }, + { + "epoch": 2.0007987220447285, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 2505 + }, + { + "epoch": 2.001597444089457, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1405, + "step": 2506 + }, + { + "epoch": 2.002396166134185, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 2507 + }, + { + "epoch": 2.0031948881789137, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2508 + }, + { + "epoch": 2.003993610223642, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1378, + "step": 2509 + }, + { + "epoch": 2.0047923322683707, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 2510 + }, + { + "epoch": 2.0055910543130993, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1415, + "step": 2511 + }, + { + "epoch": 2.0063897763578273, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 2512 + }, + { + "epoch": 2.007188498402556, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 2513 + }, + { + "epoch": 2.0079872204472844, + "grad_norm": 0.361328125, + "learning_rate": 0.0001, + "loss": 1.1317, + "step": 2514 + }, + { + "epoch": 2.008785942492013, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1367, + "step": 2515 + }, + { + "epoch": 2.009584664536741, + "grad_norm": 0.3984375, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 2516 + }, + { + "epoch": 2.0103833865814695, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1375, + "step": 2517 + }, + { + "epoch": 2.011182108626198, + "grad_norm": 0.353515625, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 2518 + }, + { + "epoch": 2.0119808306709266, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 2519 + }, + { + "epoch": 2.012779552715655, + "grad_norm": 0.412109375, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 2520 + }, + { + "epoch": 2.013578274760383, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1292, + "step": 2521 + }, + { + "epoch": 2.0143769968051117, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1409, + "step": 2522 + }, + { + "epoch": 2.0151757188498403, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 2523 + }, + { + "epoch": 2.015974440894569, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1345, + "step": 2524 + }, + { + "epoch": 2.0167731629392973, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 2525 + }, + { + "epoch": 2.0175718849840254, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 2526 + }, + { + "epoch": 2.018370607028754, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 2527 + }, + { + "epoch": 2.0191693290734825, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 2528 + }, + { + "epoch": 2.019968051118211, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1346, + "step": 2529 + }, + { + "epoch": 2.0207667731629395, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 2530 + }, + { + "epoch": 2.0215654952076676, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 2531 + }, + { + "epoch": 2.022364217252396, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2532 + }, + { + "epoch": 2.0231629392971247, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 2533 + }, + { + "epoch": 2.023961661341853, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1394, + "step": 2534 + }, + { + "epoch": 2.0247603833865813, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1388, + "step": 2535 + }, + { + "epoch": 2.02555910543131, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 2536 + }, + { + "epoch": 2.0263578274760383, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 2537 + }, + { + "epoch": 2.027156549520767, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1323, + "step": 2538 + }, + { + "epoch": 2.0279552715654954, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 2539 + }, + { + "epoch": 2.0287539936102235, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 2540 + }, + { + "epoch": 2.029552715654952, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 2541 + }, + { + "epoch": 2.0303514376996805, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 2542 + }, + { + "epoch": 2.031150159744409, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 2543 + }, + { + "epoch": 2.0319488817891376, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2544 + }, + { + "epoch": 2.0327476038338657, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 2545 + }, + { + "epoch": 2.033546325878594, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1332, + "step": 2546 + }, + { + "epoch": 2.0343450479233227, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 2547 + }, + { + "epoch": 2.0351437699680512, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 2548 + }, + { + "epoch": 2.0359424920127798, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 2549 + }, + { + "epoch": 2.036741214057508, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1366, + "step": 2550 + }, + { + "epoch": 2.0375399361022364, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 2551 + }, + { + "epoch": 2.038338658146965, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 2552 + }, + { + "epoch": 2.0391373801916934, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1327, + "step": 2553 + }, + { + "epoch": 2.0399361022364215, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1316, + "step": 2554 + }, + { + "epoch": 2.04073482428115, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 2555 + }, + { + "epoch": 2.0415335463258786, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1393, + "step": 2556 + }, + { + "epoch": 2.042332268370607, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 2557 + }, + { + "epoch": 2.0431309904153356, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 2558 + }, + { + "epoch": 2.0439297124600637, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1416, + "step": 2559 + }, + { + "epoch": 2.0447284345047922, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 2560 + }, + { + "epoch": 2.0455271565495208, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 2561 + }, + { + "epoch": 2.0463258785942493, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2562 + }, + { + "epoch": 2.047124600638978, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 2563 + }, + { + "epoch": 2.047923322683706, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1336, + "step": 2564 + }, + { + "epoch": 2.0487220447284344, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1391, + "step": 2565 + }, + { + "epoch": 2.049520766773163, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 2566 + }, + { + "epoch": 2.0503194888178915, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 2567 + }, + { + "epoch": 2.0511182108626196, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 2568 + }, + { + "epoch": 2.051916932907348, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1395, + "step": 2569 + }, + { + "epoch": 2.0527156549520766, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1393, + "step": 2570 + }, + { + "epoch": 2.053514376996805, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1332, + "step": 2571 + }, + { + "epoch": 2.0543130990415337, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 2572 + }, + { + "epoch": 2.055111821086262, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 2573 + }, + { + "epoch": 2.0559105431309903, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 2574 + }, + { + "epoch": 2.056709265175719, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2575 + }, + { + "epoch": 2.0575079872204474, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1385, + "step": 2576 + }, + { + "epoch": 2.058306709265176, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 2577 + }, + { + "epoch": 2.059105431309904, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 2578 + }, + { + "epoch": 2.0599041533546325, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1359, + "step": 2579 + }, + { + "epoch": 2.060702875399361, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 2580 + }, + { + "epoch": 2.0615015974440896, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 2581 + }, + { + "epoch": 2.062300319488818, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2582 + }, + { + "epoch": 2.063099041533546, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 2583 + }, + { + "epoch": 2.0638977635782747, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 2584 + }, + { + "epoch": 2.0646964856230032, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 2585 + }, + { + "epoch": 2.0654952076677318, + "grad_norm": 0.373046875, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 2586 + }, + { + "epoch": 2.06629392971246, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1292, + "step": 2587 + }, + { + "epoch": 2.0670926517571884, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 2588 + }, + { + "epoch": 2.067891373801917, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 2589 + }, + { + "epoch": 2.0686900958466454, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 2590 + }, + { + "epoch": 2.069488817891374, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 2591 + }, + { + "epoch": 2.070287539936102, + "grad_norm": 0.380859375, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 2592 + }, + { + "epoch": 2.0710862619808306, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 2593 + }, + { + "epoch": 2.071884984025559, + "grad_norm": 0.44921875, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 2594 + }, + { + "epoch": 2.0726837060702876, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1356, + "step": 2595 + }, + { + "epoch": 2.073482428115016, + "grad_norm": 0.39453125, + "learning_rate": 0.0001, + "loss": 1.1359, + "step": 2596 + }, + { + "epoch": 2.0742811501597442, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1395, + "step": 2597 + }, + { + "epoch": 2.0750798722044728, + "grad_norm": 0.353515625, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 2598 + }, + { + "epoch": 2.0758785942492013, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 2599 + }, + { + "epoch": 2.07667731629393, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1348, + "step": 2600 + }, + { + "epoch": 2.0774760383386583, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 2601 + }, + { + "epoch": 2.0782747603833864, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 2602 + }, + { + "epoch": 2.079073482428115, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 2603 + }, + { + "epoch": 2.0798722044728435, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1424, + "step": 2604 + }, + { + "epoch": 2.080670926517572, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 2605 + }, + { + "epoch": 2.0814696485623, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.128, + "step": 2606 + }, + { + "epoch": 2.0822683706070286, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1295, + "step": 2607 + }, + { + "epoch": 2.083067092651757, + "grad_norm": 0.34375, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 2608 + }, + { + "epoch": 2.0838658146964857, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 2609 + }, + { + "epoch": 2.084664536741214, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 2610 + }, + { + "epoch": 2.0854632587859423, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 2611 + }, + { + "epoch": 2.086261980830671, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 2612 + }, + { + "epoch": 2.0870607028753994, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 2613 + }, + { + "epoch": 2.087859424920128, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1321, + "step": 2614 + }, + { + "epoch": 2.0886581469648564, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1345, + "step": 2615 + }, + { + "epoch": 2.0894568690095845, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 2616 + }, + { + "epoch": 2.090255591054313, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 2617 + }, + { + "epoch": 2.0910543130990416, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 2618 + }, + { + "epoch": 2.09185303514377, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1337, + "step": 2619 + }, + { + "epoch": 2.0926517571884986, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 2620 + }, + { + "epoch": 2.0934504792332267, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 2621 + }, + { + "epoch": 2.094249201277955, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1345, + "step": 2622 + }, + { + "epoch": 2.0950479233226837, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 2623 + }, + { + "epoch": 2.0958466453674123, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 2624 + }, + { + "epoch": 2.0966453674121404, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1349, + "step": 2625 + }, + { + "epoch": 2.097444089456869, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 2626 + }, + { + "epoch": 2.0982428115015974, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 2627 + }, + { + "epoch": 2.099041533546326, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.133, + "step": 2628 + }, + { + "epoch": 2.0998402555910545, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 2629 + }, + { + "epoch": 2.1006389776357826, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 2630 + }, + { + "epoch": 2.101437699680511, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1301, + "step": 2631 + }, + { + "epoch": 2.1022364217252396, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1357, + "step": 2632 + }, + { + "epoch": 2.103035143769968, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 2633 + }, + { + "epoch": 2.1038338658146967, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 2634 + }, + { + "epoch": 2.1046325878594248, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 2635 + }, + { + "epoch": 2.1054313099041533, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 2636 + }, + { + "epoch": 2.106230031948882, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 2637 + }, + { + "epoch": 2.1070287539936103, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1348, + "step": 2638 + }, + { + "epoch": 2.107827476038339, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 2639 + }, + { + "epoch": 2.108626198083067, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 2640 + }, + { + "epoch": 2.1094249201277955, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 2641 + }, + { + "epoch": 2.110223642172524, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1319, + "step": 2642 + }, + { + "epoch": 2.1110223642172525, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1325, + "step": 2643 + }, + { + "epoch": 2.1118210862619806, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1333, + "step": 2644 + }, + { + "epoch": 2.112619808306709, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 2645 + }, + { + "epoch": 2.1134185303514377, + "grad_norm": 0.365234375, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2646 + }, + { + "epoch": 2.114217252396166, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 2647 + }, + { + "epoch": 2.1150159744408947, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 2648 + }, + { + "epoch": 2.115814696485623, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1369, + "step": 2649 + }, + { + "epoch": 2.1166134185303513, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 2650 + }, + { + "epoch": 2.11741214057508, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.135, + "step": 2651 + }, + { + "epoch": 2.1182108626198084, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1287, + "step": 2652 + }, + { + "epoch": 2.119009584664537, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1317, + "step": 2653 + }, + { + "epoch": 2.119808306709265, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 2654 + }, + { + "epoch": 2.1206070287539935, + "grad_norm": 0.36328125, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 2655 + }, + { + "epoch": 2.121405750798722, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1428, + "step": 2656 + }, + { + "epoch": 2.1222044728434506, + "grad_norm": 0.3984375, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 2657 + }, + { + "epoch": 2.123003194888179, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 2658 + }, + { + "epoch": 2.123801916932907, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 2659 + }, + { + "epoch": 2.1246006389776357, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 2660 + }, + { + "epoch": 2.1253993610223643, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 2661 + }, + { + "epoch": 2.126198083067093, + "grad_norm": 0.47265625, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 2662 + }, + { + "epoch": 2.126996805111821, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 2663 + }, + { + "epoch": 2.1277955271565494, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 2664 + }, + { + "epoch": 2.128594249201278, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 2665 + }, + { + "epoch": 2.1293929712460065, + "grad_norm": 0.357421875, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 2666 + }, + { + "epoch": 2.130191693290735, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1374, + "step": 2667 + }, + { + "epoch": 2.130990415335463, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1383, + "step": 2668 + }, + { + "epoch": 2.1317891373801916, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1356, + "step": 2669 + }, + { + "epoch": 2.13258785942492, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 2670 + }, + { + "epoch": 2.1333865814696487, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2671 + }, + { + "epoch": 2.134185303514377, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 2672 + }, + { + "epoch": 2.1349840255591053, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 2673 + }, + { + "epoch": 2.135782747603834, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 2674 + }, + { + "epoch": 2.1365814696485623, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 2675 + }, + { + "epoch": 2.137380191693291, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 2676 + }, + { + "epoch": 2.1381789137380194, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 2677 + }, + { + "epoch": 2.1389776357827475, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.128, + "step": 2678 + }, + { + "epoch": 2.139776357827476, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 2679 + }, + { + "epoch": 2.1405750798722045, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 2680 + }, + { + "epoch": 2.141373801916933, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2681 + }, + { + "epoch": 2.142172523961661, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 2682 + }, + { + "epoch": 2.1429712460063897, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 2683 + }, + { + "epoch": 2.143769968051118, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1447, + "step": 2684 + }, + { + "epoch": 2.1445686900958467, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 2685 + }, + { + "epoch": 2.1453674121405752, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 2686 + }, + { + "epoch": 2.1461661341853033, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 2687 + }, + { + "epoch": 2.146964856230032, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.1311, + "step": 2688 + }, + { + "epoch": 2.1477635782747604, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1294, + "step": 2689 + }, + { + "epoch": 2.148562300319489, + "grad_norm": 0.361328125, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 2690 + }, + { + "epoch": 2.1493610223642174, + "grad_norm": 0.50390625, + "learning_rate": 0.0001, + "loss": 1.1332, + "step": 2691 + }, + { + "epoch": 2.1501597444089455, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 2692 + }, + { + "epoch": 2.150958466453674, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 2693 + }, + { + "epoch": 2.1517571884984026, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 2694 + }, + { + "epoch": 2.152555910543131, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 2695 + }, + { + "epoch": 2.1533546325878596, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 2696 + }, + { + "epoch": 2.1541533546325877, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 2697 + }, + { + "epoch": 2.1549520766773163, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 2698 + }, + { + "epoch": 2.155750798722045, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 2699 + }, + { + "epoch": 2.1565495207667733, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 2700 + }, + { + "epoch": 2.1573482428115014, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 2701 + }, + { + "epoch": 2.15814696485623, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1305, + "step": 2702 + }, + { + "epoch": 2.1589456869009584, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 2703 + }, + { + "epoch": 2.159744408945687, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 2704 + }, + { + "epoch": 2.1605431309904155, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 2705 + }, + { + "epoch": 2.1613418530351436, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 2706 + }, + { + "epoch": 2.162140575079872, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 2707 + }, + { + "epoch": 2.1629392971246006, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 2708 + }, + { + "epoch": 2.163738019169329, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 2709 + }, + { + "epoch": 2.1645367412140577, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 2710 + }, + { + "epoch": 2.165335463258786, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 2711 + }, + { + "epoch": 2.1661341853035143, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 2712 + }, + { + "epoch": 2.166932907348243, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 2713 + }, + { + "epoch": 2.1677316293929714, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 2714 + }, + { + "epoch": 2.1685303514377, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1334, + "step": 2715 + }, + { + "epoch": 2.169329073482428, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1367, + "step": 2716 + }, + { + "epoch": 2.1701277955271565, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1399, + "step": 2717 + }, + { + "epoch": 2.170926517571885, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1384, + "step": 2718 + }, + { + "epoch": 2.1717252396166136, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 2719 + }, + { + "epoch": 2.1725239616613417, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1347, + "step": 2720 + }, + { + "epoch": 2.17332268370607, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 2721 + }, + { + "epoch": 2.1741214057507987, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 2722 + }, + { + "epoch": 2.1749201277955272, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 2723 + }, + { + "epoch": 2.1757188498402558, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 2724 + }, + { + "epoch": 2.176517571884984, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 2725 + }, + { + "epoch": 2.1773162939297124, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 2726 + }, + { + "epoch": 2.178115015974441, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 2727 + }, + { + "epoch": 2.1789137380191694, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 2728 + }, + { + "epoch": 2.179712460063898, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 2729 + }, + { + "epoch": 2.180511182108626, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 2730 + }, + { + "epoch": 2.1813099041533546, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 2731 + }, + { + "epoch": 2.182108626198083, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 2732 + }, + { + "epoch": 2.1829073482428116, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1347, + "step": 2733 + }, + { + "epoch": 2.18370607028754, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 2734 + }, + { + "epoch": 2.1845047923322682, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 2735 + }, + { + "epoch": 2.1853035143769968, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 2736 + }, + { + "epoch": 2.1861022364217253, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.134, + "step": 2737 + }, + { + "epoch": 2.186900958466454, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 2738 + }, + { + "epoch": 2.187699680511182, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 2739 + }, + { + "epoch": 2.1884984025559104, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 2740 + }, + { + "epoch": 2.189297124600639, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 2741 + }, + { + "epoch": 2.1900958466453675, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1382, + "step": 2742 + }, + { + "epoch": 2.190894568690096, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 2743 + }, + { + "epoch": 2.191693290734824, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 2744 + }, + { + "epoch": 2.1924920127795526, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 2745 + }, + { + "epoch": 2.193290734824281, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1396, + "step": 2746 + }, + { + "epoch": 2.1940894568690097, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 2747 + }, + { + "epoch": 2.194888178913738, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 2748 + }, + { + "epoch": 2.1956869009584663, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 2749 + }, + { + "epoch": 2.196485623003195, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 2750 + }, + { + "epoch": 2.1972843450479234, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1335, + "step": 2751 + }, + { + "epoch": 2.198083067092652, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 2752 + }, + { + "epoch": 2.1988817891373804, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 2753 + }, + { + "epoch": 2.1996805111821085, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 2754 + }, + { + "epoch": 2.200479233226837, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1349, + "step": 2755 + }, + { + "epoch": 2.2012779552715656, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1356, + "step": 2756 + }, + { + "epoch": 2.202076677316294, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1292, + "step": 2757 + }, + { + "epoch": 2.202875399361022, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 2758 + }, + { + "epoch": 2.2036741214057507, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 2759 + }, + { + "epoch": 2.2044728434504792, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1323, + "step": 2760 + }, + { + "epoch": 2.2052715654952078, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 2761 + }, + { + "epoch": 2.2060702875399363, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1311, + "step": 2762 + }, + { + "epoch": 2.2068690095846644, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 2763 + }, + { + "epoch": 2.207667731629393, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1347, + "step": 2764 + }, + { + "epoch": 2.2084664536741214, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 2765 + }, + { + "epoch": 2.20926517571885, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 2766 + }, + { + "epoch": 2.2100638977635785, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2767 + }, + { + "epoch": 2.2108626198083066, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 2768 + }, + { + "epoch": 2.211661341853035, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 2769 + }, + { + "epoch": 2.2124600638977636, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1316, + "step": 2770 + }, + { + "epoch": 2.213258785942492, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 2771 + }, + { + "epoch": 2.2140575079872207, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1283, + "step": 2772 + }, + { + "epoch": 2.2148562300319488, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 2773 + }, + { + "epoch": 2.2156549520766773, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 2774 + }, + { + "epoch": 2.216453674121406, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 2775 + }, + { + "epoch": 2.2172523961661343, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 2776 + }, + { + "epoch": 2.2180511182108624, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1402, + "step": 2777 + }, + { + "epoch": 2.218849840255591, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.139, + "step": 2778 + }, + { + "epoch": 2.2196485623003195, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 2779 + }, + { + "epoch": 2.220447284345048, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 2780 + }, + { + "epoch": 2.2212460063897765, + "grad_norm": 0.392578125, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 2781 + }, + { + "epoch": 2.2220447284345046, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1295, + "step": 2782 + }, + { + "epoch": 2.222843450479233, + "grad_norm": 0.359375, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 2783 + }, + { + "epoch": 2.2236421725239617, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 2784 + }, + { + "epoch": 2.22444089456869, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.1333, + "step": 2785 + }, + { + "epoch": 2.2252396166134187, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1337, + "step": 2786 + }, + { + "epoch": 2.226038338658147, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 2787 + }, + { + "epoch": 2.2268370607028753, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 2788 + }, + { + "epoch": 2.227635782747604, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 2789 + }, + { + "epoch": 2.2284345047923324, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1346, + "step": 2790 + }, + { + "epoch": 2.229233226837061, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1317, + "step": 2791 + }, + { + "epoch": 2.230031948881789, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1301, + "step": 2792 + }, + { + "epoch": 2.2308306709265175, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 2793 + }, + { + "epoch": 2.231629392971246, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1319, + "step": 2794 + }, + { + "epoch": 2.2324281150159746, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 2795 + }, + { + "epoch": 2.2332268370607027, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 2796 + }, + { + "epoch": 2.234025559105431, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 2797 + }, + { + "epoch": 2.2348242811501597, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1346, + "step": 2798 + }, + { + "epoch": 2.2356230031948883, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 2799 + }, + { + "epoch": 2.236421725239617, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 2800 + }, + { + "epoch": 2.237220447284345, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 2801 + }, + { + "epoch": 2.2380191693290734, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 2802 + }, + { + "epoch": 2.238817891373802, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 2803 + }, + { + "epoch": 2.2396166134185305, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 2804 + }, + { + "epoch": 2.2404153354632586, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1292, + "step": 2805 + }, + { + "epoch": 2.241214057507987, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 2806 + }, + { + "epoch": 2.2420127795527156, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1329, + "step": 2807 + }, + { + "epoch": 2.242811501597444, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 2808 + }, + { + "epoch": 2.2436102236421727, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 2809 + }, + { + "epoch": 2.244408945686901, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 2810 + }, + { + "epoch": 2.2452076677316293, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 2811 + }, + { + "epoch": 2.246006389776358, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 2812 + }, + { + "epoch": 2.2468051118210863, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 2813 + }, + { + "epoch": 2.247603833865815, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 2814 + }, + { + "epoch": 2.248402555910543, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 2815 + }, + { + "epoch": 2.2492012779552715, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 2816 + }, + { + "epoch": 2.25, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 2817 + }, + { + "epoch": 2.2507987220447285, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1316, + "step": 2818 + }, + { + "epoch": 2.251597444089457, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 2819 + }, + { + "epoch": 2.252396166134185, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 2820 + }, + { + "epoch": 2.2531948881789137, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 2821 + }, + { + "epoch": 2.253993610223642, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.134, + "step": 2822 + }, + { + "epoch": 2.2547923322683707, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 2823 + }, + { + "epoch": 2.255591054313099, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 2824 + }, + { + "epoch": 2.2563897763578273, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2825 + }, + { + "epoch": 2.257188498402556, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 2826 + }, + { + "epoch": 2.2579872204472844, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1401, + "step": 2827 + }, + { + "epoch": 2.258785942492013, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 2828 + }, + { + "epoch": 2.2595846645367414, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 2829 + }, + { + "epoch": 2.2603833865814695, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1269, + "step": 2830 + }, + { + "epoch": 2.261182108626198, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 2831 + }, + { + "epoch": 2.2619808306709266, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 2832 + }, + { + "epoch": 2.262779552715655, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 2833 + }, + { + "epoch": 2.263578274760383, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 2834 + }, + { + "epoch": 2.2643769968051117, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 2835 + }, + { + "epoch": 2.2651757188498403, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 2836 + }, + { + "epoch": 2.265974440894569, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 2837 + }, + { + "epoch": 2.2667731629392973, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 2838 + }, + { + "epoch": 2.2675718849840254, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 2839 + }, + { + "epoch": 2.268370607028754, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1344, + "step": 2840 + }, + { + "epoch": 2.2691693290734825, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1334, + "step": 2841 + }, + { + "epoch": 2.269968051118211, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 2842 + }, + { + "epoch": 2.270766773162939, + "grad_norm": 0.365234375, + "learning_rate": 0.0001, + "loss": 1.1283, + "step": 2843 + }, + { + "epoch": 2.2715654952076676, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 2844 + }, + { + "epoch": 2.272364217252396, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1362, + "step": 2845 + }, + { + "epoch": 2.2731629392971247, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 2846 + }, + { + "epoch": 2.273961661341853, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 2847 + }, + { + "epoch": 2.2747603833865817, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1321, + "step": 2848 + }, + { + "epoch": 2.27555910543131, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 2849 + }, + { + "epoch": 2.2763578274760383, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 2850 + }, + { + "epoch": 2.277156549520767, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1374, + "step": 2851 + }, + { + "epoch": 2.2779552715654954, + "grad_norm": 0.359375, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 2852 + }, + { + "epoch": 2.2787539936102235, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1368, + "step": 2853 + }, + { + "epoch": 2.279552715654952, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 2854 + }, + { + "epoch": 2.2803514376996805, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 2855 + }, + { + "epoch": 2.281150159744409, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 2856 + }, + { + "epoch": 2.2819488817891376, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 2857 + }, + { + "epoch": 2.2827476038338657, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 2858 + }, + { + "epoch": 2.283546325878594, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1421, + "step": 2859 + }, + { + "epoch": 2.2843450479233227, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2860 + }, + { + "epoch": 2.2851437699680512, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1302, + "step": 2861 + }, + { + "epoch": 2.2859424920127793, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 2862 + }, + { + "epoch": 2.286741214057508, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 2863 + }, + { + "epoch": 2.2875399361022364, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1362, + "step": 2864 + }, + { + "epoch": 2.288338658146965, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1345, + "step": 2865 + }, + { + "epoch": 2.2891373801916934, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 2866 + }, + { + "epoch": 2.289936102236422, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1373, + "step": 2867 + }, + { + "epoch": 2.29073482428115, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 2868 + }, + { + "epoch": 2.2915335463258786, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 2869 + }, + { + "epoch": 2.292332268370607, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 2870 + }, + { + "epoch": 2.2931309904153356, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 2871 + }, + { + "epoch": 2.2939297124600637, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 2872 + }, + { + "epoch": 2.2947284345047922, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 2873 + }, + { + "epoch": 2.2955271565495208, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.128, + "step": 2874 + }, + { + "epoch": 2.2963258785942493, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 2875 + }, + { + "epoch": 2.297124600638978, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 2876 + }, + { + "epoch": 2.297923322683706, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 2877 + }, + { + "epoch": 2.2987220447284344, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 2878 + }, + { + "epoch": 2.299520766773163, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 2879 + }, + { + "epoch": 2.3003194888178915, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 2880 + }, + { + "epoch": 2.3011182108626196, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 2881 + }, + { + "epoch": 2.301916932907348, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1452, + "step": 2882 + }, + { + "epoch": 2.3027156549520766, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 2883 + }, + { + "epoch": 2.303514376996805, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1329, + "step": 2884 + }, + { + "epoch": 2.3043130990415337, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 2885 + }, + { + "epoch": 2.3051118210862622, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 2886 + }, + { + "epoch": 2.3059105431309903, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1333, + "step": 2887 + }, + { + "epoch": 2.306709265175719, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 2888 + }, + { + "epoch": 2.3075079872204474, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1335, + "step": 2889 + }, + { + "epoch": 2.308306709265176, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 2890 + }, + { + "epoch": 2.309105431309904, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 2891 + }, + { + "epoch": 2.3099041533546325, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 2892 + }, + { + "epoch": 2.310702875399361, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 2893 + }, + { + "epoch": 2.3115015974440896, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1352, + "step": 2894 + }, + { + "epoch": 2.312300319488818, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 2895 + }, + { + "epoch": 2.313099041533546, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 2896 + }, + { + "epoch": 2.3138977635782747, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 2897 + }, + { + "epoch": 2.3146964856230032, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 2898 + }, + { + "epoch": 2.3154952076677318, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1329, + "step": 2899 + }, + { + "epoch": 2.31629392971246, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1351, + "step": 2900 + }, + { + "epoch": 2.3170926517571884, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 2901 + }, + { + "epoch": 2.317891373801917, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 2902 + }, + { + "epoch": 2.3186900958466454, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 2903 + }, + { + "epoch": 2.319488817891374, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 2904 + }, + { + "epoch": 2.3202875399361025, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1369, + "step": 2905 + }, + { + "epoch": 2.3210862619808306, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 2906 + }, + { + "epoch": 2.321884984025559, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 2907 + }, + { + "epoch": 2.3226837060702876, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 2908 + }, + { + "epoch": 2.323482428115016, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2909 + }, + { + "epoch": 2.3242811501597442, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 2910 + }, + { + "epoch": 2.3250798722044728, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 2911 + }, + { + "epoch": 2.3258785942492013, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 2912 + }, + { + "epoch": 2.32667731629393, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1334, + "step": 2913 + }, + { + "epoch": 2.3274760383386583, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 2914 + }, + { + "epoch": 2.3282747603833864, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 2915 + }, + { + "epoch": 2.329073482428115, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 2916 + }, + { + "epoch": 2.3298722044728435, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 2917 + }, + { + "epoch": 2.330670926517572, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 2918 + }, + { + "epoch": 2.3314696485623, + "grad_norm": 0.359375, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 2919 + }, + { + "epoch": 2.3322683706070286, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 2920 + }, + { + "epoch": 2.333067092651757, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 2921 + }, + { + "epoch": 2.3338658146964857, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 2922 + }, + { + "epoch": 2.334664536741214, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 2923 + }, + { + "epoch": 2.3354632587859427, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 2924 + }, + { + "epoch": 2.336261980830671, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1335, + "step": 2925 + }, + { + "epoch": 2.3370607028753994, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1347, + "step": 2926 + }, + { + "epoch": 2.337859424920128, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1354, + "step": 2927 + }, + { + "epoch": 2.3386581469648564, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 2928 + }, + { + "epoch": 2.3394568690095845, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 2929 + }, + { + "epoch": 2.340255591054313, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 2930 + }, + { + "epoch": 2.3410543130990416, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 2931 + }, + { + "epoch": 2.34185303514377, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 2932 + }, + { + "epoch": 2.3426517571884986, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 2933 + }, + { + "epoch": 2.3434504792332267, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 2934 + }, + { + "epoch": 2.344249201277955, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 2935 + }, + { + "epoch": 2.3450479233226837, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 2936 + }, + { + "epoch": 2.3458466453674123, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 2937 + }, + { + "epoch": 2.3466453674121404, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1349, + "step": 2938 + }, + { + "epoch": 2.347444089456869, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 2939 + }, + { + "epoch": 2.3482428115015974, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 2940 + }, + { + "epoch": 2.349041533546326, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1283, + "step": 2941 + }, + { + "epoch": 2.3498402555910545, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.134, + "step": 2942 + }, + { + "epoch": 2.3506389776357826, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 2943 + }, + { + "epoch": 2.351437699680511, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1403, + "step": 2944 + }, + { + "epoch": 2.3522364217252396, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 2945 + }, + { + "epoch": 2.353035143769968, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 2946 + }, + { + "epoch": 2.3538338658146967, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 2947 + }, + { + "epoch": 2.3546325878594248, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1311, + "step": 2948 + }, + { + "epoch": 2.3554313099041533, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 2949 + }, + { + "epoch": 2.356230031948882, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 2950 + }, + { + "epoch": 2.3570287539936103, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 2951 + }, + { + "epoch": 2.357827476038339, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 2952 + }, + { + "epoch": 2.358626198083067, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 2953 + }, + { + "epoch": 2.3594249201277955, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 2954 + }, + { + "epoch": 2.360223642172524, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 2955 + }, + { + "epoch": 2.3610223642172525, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 2956 + }, + { + "epoch": 2.3618210862619806, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 2957 + }, + { + "epoch": 2.362619808306709, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 2958 + }, + { + "epoch": 2.3634185303514377, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 2959 + }, + { + "epoch": 2.364217252396166, + "grad_norm": 0.357421875, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 2960 + }, + { + "epoch": 2.3650159744408947, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 2961 + }, + { + "epoch": 2.365814696485623, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 2962 + }, + { + "epoch": 2.3666134185303513, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 2963 + }, + { + "epoch": 2.36741214057508, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 2964 + }, + { + "epoch": 2.3682108626198084, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1323, + "step": 2965 + }, + { + "epoch": 2.369009584664537, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 2966 + }, + { + "epoch": 2.369808306709265, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 2967 + }, + { + "epoch": 2.3706070287539935, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 2968 + }, + { + "epoch": 2.371405750798722, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 2969 + }, + { + "epoch": 2.3722044728434506, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 2970 + }, + { + "epoch": 2.373003194888179, + "grad_norm": 0.328125, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 2971 + }, + { + "epoch": 2.373801916932907, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 2972 + }, + { + "epoch": 2.3746006389776357, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1362, + "step": 2973 + }, + { + "epoch": 2.3753993610223643, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1348, + "step": 2974 + }, + { + "epoch": 2.376198083067093, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 2975 + }, + { + "epoch": 2.376996805111821, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 2976 + }, + { + "epoch": 2.3777955271565494, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 2977 + }, + { + "epoch": 2.378594249201278, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 2978 + }, + { + "epoch": 2.3793929712460065, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 2979 + }, + { + "epoch": 2.380191693290735, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 2980 + }, + { + "epoch": 2.380990415335463, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 2981 + }, + { + "epoch": 2.3817891373801916, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 2982 + }, + { + "epoch": 2.38258785942492, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 2983 + }, + { + "epoch": 2.3833865814696487, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 2984 + }, + { + "epoch": 2.384185303514377, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 2985 + }, + { + "epoch": 2.3849840255591053, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 2986 + }, + { + "epoch": 2.385782747603834, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 2987 + }, + { + "epoch": 2.3865814696485623, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 2988 + }, + { + "epoch": 2.387380191693291, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 2989 + }, + { + "epoch": 2.3881789137380194, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 2990 + }, + { + "epoch": 2.3889776357827475, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 2991 + }, + { + "epoch": 2.389776357827476, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 2992 + }, + { + "epoch": 2.3905750798722045, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 2993 + }, + { + "epoch": 2.391373801916933, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 2994 + }, + { + "epoch": 2.392172523961661, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 2995 + }, + { + "epoch": 2.3929712460063897, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 2996 + }, + { + "epoch": 2.393769968051118, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 2997 + }, + { + "epoch": 2.3945686900958467, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1295, + "step": 2998 + }, + { + "epoch": 2.3953674121405752, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1356, + "step": 2999 + }, + { + "epoch": 2.3961661341853033, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 3000 + }, + { + "epoch": 2.396964856230032, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 3001 + }, + { + "epoch": 2.3977635782747604, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 3002 + }, + { + "epoch": 2.398562300319489, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 3003 + }, + { + "epoch": 2.3993610223642174, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3004 + }, + { + "epoch": 2.4001597444089455, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 3005 + }, + { + "epoch": 2.400958466453674, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3006 + }, + { + "epoch": 2.4017571884984026, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 3007 + }, + { + "epoch": 2.402555910543131, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 3008 + }, + { + "epoch": 2.4033546325878596, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1248, + "step": 3009 + }, + { + "epoch": 2.4041533546325877, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1311, + "step": 3010 + }, + { + "epoch": 2.4049520766773163, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 3011 + }, + { + "epoch": 2.405750798722045, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 3012 + }, + { + "epoch": 2.4065495207667733, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 3013 + }, + { + "epoch": 2.4073482428115014, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 3014 + }, + { + "epoch": 2.40814696485623, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1316, + "step": 3015 + }, + { + "epoch": 2.4089456869009584, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1301, + "step": 3016 + }, + { + "epoch": 2.409744408945687, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 3017 + }, + { + "epoch": 2.4105431309904155, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 3018 + }, + { + "epoch": 2.4113418530351436, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 3019 + }, + { + "epoch": 2.412140575079872, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 3020 + }, + { + "epoch": 2.4129392971246006, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1369, + "step": 3021 + }, + { + "epoch": 2.413738019169329, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 3022 + }, + { + "epoch": 2.4145367412140573, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 3023 + }, + { + "epoch": 2.415335463258786, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1302, + "step": 3024 + }, + { + "epoch": 2.4161341853035143, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 3025 + }, + { + "epoch": 2.416932907348243, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 3026 + }, + { + "epoch": 2.4177316293929714, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 3027 + }, + { + "epoch": 2.4185303514377, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 3028 + }, + { + "epoch": 2.419329073482428, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 3029 + }, + { + "epoch": 2.4201277955271565, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 3030 + }, + { + "epoch": 2.420926517571885, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1346, + "step": 3031 + }, + { + "epoch": 2.4217252396166136, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3032 + }, + { + "epoch": 2.4225239616613417, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 3033 + }, + { + "epoch": 2.42332268370607, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 3034 + }, + { + "epoch": 2.4241214057507987, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 3035 + }, + { + "epoch": 2.4249201277955272, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1273, + "step": 3036 + }, + { + "epoch": 2.4257188498402558, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 3037 + }, + { + "epoch": 2.426517571884984, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 3038 + }, + { + "epoch": 2.4273162939297124, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 3039 + }, + { + "epoch": 2.428115015974441, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 3040 + }, + { + "epoch": 2.4289137380191694, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.135, + "step": 3041 + }, + { + "epoch": 2.4297124600638975, + "grad_norm": 0.357421875, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 3042 + }, + { + "epoch": 2.430511182108626, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 3043 + }, + { + "epoch": 2.4313099041533546, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1329, + "step": 3044 + }, + { + "epoch": 2.432108626198083, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1294, + "step": 3045 + }, + { + "epoch": 2.4329073482428116, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1328, + "step": 3046 + }, + { + "epoch": 2.43370607028754, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 3047 + }, + { + "epoch": 2.4345047923322682, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 3048 + }, + { + "epoch": 2.4353035143769968, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 3049 + }, + { + "epoch": 2.4361022364217253, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 3050 + }, + { + "epoch": 2.436900958466454, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 3051 + }, + { + "epoch": 2.437699680511182, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 3052 + }, + { + "epoch": 2.4384984025559104, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1235, + "step": 3053 + }, + { + "epoch": 2.439297124600639, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 3054 + }, + { + "epoch": 2.4400958466453675, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 3055 + }, + { + "epoch": 2.440894568690096, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 3056 + }, + { + "epoch": 2.441693290734824, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 3057 + }, + { + "epoch": 2.4424920127795526, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.127, + "step": 3058 + }, + { + "epoch": 2.443290734824281, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 3059 + }, + { + "epoch": 2.4440894568690097, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 3060 + }, + { + "epoch": 2.4448881789137378, + "grad_norm": 0.39453125, + "learning_rate": 0.0001, + "loss": 1.1321, + "step": 3061 + }, + { + "epoch": 2.4456869009584663, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 3062 + }, + { + "epoch": 2.446485623003195, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 3063 + }, + { + "epoch": 2.4472843450479234, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1335, + "step": 3064 + }, + { + "epoch": 2.448083067092652, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1321, + "step": 3065 + }, + { + "epoch": 2.4488817891373804, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3066 + }, + { + "epoch": 2.4496805111821085, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3067 + }, + { + "epoch": 2.450479233226837, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 3068 + }, + { + "epoch": 2.4512779552715656, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 3069 + }, + { + "epoch": 2.452076677316294, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 3070 + }, + { + "epoch": 2.452875399361022, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 3071 + }, + { + "epoch": 2.4536741214057507, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 3072 + }, + { + "epoch": 2.4544728434504792, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1366, + "step": 3073 + }, + { + "epoch": 2.4552715654952078, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 3074 + }, + { + "epoch": 2.4560702875399363, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 3075 + }, + { + "epoch": 2.4568690095846644, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 3076 + }, + { + "epoch": 2.457667731629393, + "grad_norm": 0.37890625, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 3077 + }, + { + "epoch": 2.4584664536741214, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 3078 + }, + { + "epoch": 2.45926517571885, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 3079 + }, + { + "epoch": 2.460063897763578, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3080 + }, + { + "epoch": 2.4608626198083066, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 3081 + }, + { + "epoch": 2.461661341853035, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 3082 + }, + { + "epoch": 2.4624600638977636, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 3083 + }, + { + "epoch": 2.463258785942492, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 3084 + }, + { + "epoch": 2.4640575079872207, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 3085 + }, + { + "epoch": 2.4648562300319488, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1302, + "step": 3086 + }, + { + "epoch": 2.4656549520766773, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 3087 + }, + { + "epoch": 2.466453674121406, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 3088 + }, + { + "epoch": 2.4672523961661343, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1337, + "step": 3089 + }, + { + "epoch": 2.4680511182108624, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 3090 + }, + { + "epoch": 2.468849840255591, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 3091 + }, + { + "epoch": 2.4696485623003195, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 3092 + }, + { + "epoch": 2.470447284345048, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 3093 + }, + { + "epoch": 2.4712460063897765, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 3094 + }, + { + "epoch": 2.4720447284345046, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3095 + }, + { + "epoch": 2.472843450479233, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1328, + "step": 3096 + }, + { + "epoch": 2.4736421725239617, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 3097 + }, + { + "epoch": 2.47444089456869, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1362, + "step": 3098 + }, + { + "epoch": 2.4752396166134183, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 3099 + }, + { + "epoch": 2.476038338658147, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 3100 + }, + { + "epoch": 2.4768370607028753, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 3101 + }, + { + "epoch": 2.477635782747604, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 3102 + }, + { + "epoch": 2.4784345047923324, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1359, + "step": 3103 + }, + { + "epoch": 2.479233226837061, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3104 + }, + { + "epoch": 2.480031948881789, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 3105 + }, + { + "epoch": 2.4808306709265175, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 3106 + }, + { + "epoch": 2.481629392971246, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 3107 + }, + { + "epoch": 2.4824281150159746, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 3108 + }, + { + "epoch": 2.4832268370607027, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 3109 + }, + { + "epoch": 2.484025559105431, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 3110 + }, + { + "epoch": 2.4848242811501597, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 3111 + }, + { + "epoch": 2.4856230031948883, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 3112 + }, + { + "epoch": 2.486421725239617, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 3113 + }, + { + "epoch": 2.487220447284345, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 3114 + }, + { + "epoch": 2.4880191693290734, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 3115 + }, + { + "epoch": 2.488817891373802, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 3116 + }, + { + "epoch": 2.4896166134185305, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1316, + "step": 3117 + }, + { + "epoch": 2.4904153354632586, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 3118 + }, + { + "epoch": 2.491214057507987, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 3119 + }, + { + "epoch": 2.4920127795527156, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 3120 + }, + { + "epoch": 2.492811501597444, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 3121 + }, + { + "epoch": 2.4936102236421727, + "grad_norm": 0.375, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 3122 + }, + { + "epoch": 2.494408945686901, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 3123 + }, + { + "epoch": 2.4952076677316293, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 3124 + }, + { + "epoch": 2.496006389776358, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 3125 + }, + { + "epoch": 2.4968051118210863, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 3126 + }, + { + "epoch": 2.497603833865815, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 3127 + }, + { + "epoch": 2.498402555910543, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 3128 + }, + { + "epoch": 2.4992012779552715, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 3129 + }, + { + "epoch": 2.5, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 3130 + }, + { + "epoch": 2.5007987220447285, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 3131 + }, + { + "epoch": 2.501597444089457, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1292, + "step": 3132 + }, + { + "epoch": 2.502396166134185, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.127, + "step": 3133 + }, + { + "epoch": 2.5031948881789137, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3134 + }, + { + "epoch": 2.503993610223642, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 3135 + }, + { + "epoch": 2.5047923322683707, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1347, + "step": 3136 + }, + { + "epoch": 2.505591054313099, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 3137 + }, + { + "epoch": 2.5063897763578273, + "grad_norm": 0.328125, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 3138 + }, + { + "epoch": 2.507188498402556, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 3139 + }, + { + "epoch": 2.5079872204472844, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 3140 + }, + { + "epoch": 2.508785942492013, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 3141 + }, + { + "epoch": 2.5095846645367414, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1292, + "step": 3142 + }, + { + "epoch": 2.5103833865814695, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 3143 + }, + { + "epoch": 2.511182108626198, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 3144 + }, + { + "epoch": 2.5119808306709266, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 3145 + }, + { + "epoch": 2.512779552715655, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 3146 + }, + { + "epoch": 2.513578274760383, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 3147 + }, + { + "epoch": 2.5143769968051117, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 3148 + }, + { + "epoch": 2.5151757188498403, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1404, + "step": 3149 + }, + { + "epoch": 2.515974440894569, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.133, + "step": 3150 + }, + { + "epoch": 2.5167731629392973, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 3151 + }, + { + "epoch": 2.5175718849840254, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 3152 + }, + { + "epoch": 2.518370607028754, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 3153 + }, + { + "epoch": 2.5191693290734825, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 3154 + }, + { + "epoch": 2.519968051118211, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 3155 + }, + { + "epoch": 2.520766773162939, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1372, + "step": 3156 + }, + { + "epoch": 2.5215654952076676, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.133, + "step": 3157 + }, + { + "epoch": 2.522364217252396, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 3158 + }, + { + "epoch": 2.5231629392971247, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 3159 + }, + { + "epoch": 2.523961661341853, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 3160 + }, + { + "epoch": 2.5247603833865817, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 3161 + }, + { + "epoch": 2.52555910543131, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 3162 + }, + { + "epoch": 2.5263578274760383, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1269, + "step": 3163 + }, + { + "epoch": 2.527156549520767, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 3164 + }, + { + "epoch": 2.527955271565495, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1361, + "step": 3165 + }, + { + "epoch": 2.5287539936102235, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.136, + "step": 3166 + }, + { + "epoch": 2.529552715654952, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 3167 + }, + { + "epoch": 2.5303514376996805, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 3168 + }, + { + "epoch": 2.531150159744409, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 3169 + }, + { + "epoch": 2.5319488817891376, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 3170 + }, + { + "epoch": 2.5327476038338657, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 3171 + }, + { + "epoch": 2.533546325878594, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 3172 + }, + { + "epoch": 2.5343450479233227, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 3173 + }, + { + "epoch": 2.5351437699680512, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 3174 + }, + { + "epoch": 2.5359424920127793, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 3175 + }, + { + "epoch": 2.536741214057508, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 3176 + }, + { + "epoch": 2.5375399361022364, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 3177 + }, + { + "epoch": 2.538338658146965, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.137, + "step": 3178 + }, + { + "epoch": 2.5391373801916934, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.133, + "step": 3179 + }, + { + "epoch": 2.539936102236422, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1316, + "step": 3180 + }, + { + "epoch": 2.54073482428115, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 3181 + }, + { + "epoch": 2.5415335463258786, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 3182 + }, + { + "epoch": 2.542332268370607, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 3183 + }, + { + "epoch": 2.543130990415335, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 3184 + }, + { + "epoch": 2.5439297124600637, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1329, + "step": 3185 + }, + { + "epoch": 2.5447284345047922, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1269, + "step": 3186 + }, + { + "epoch": 2.5455271565495208, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 3187 + }, + { + "epoch": 2.5463258785942493, + "grad_norm": 0.40234375, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 3188 + }, + { + "epoch": 2.547124600638978, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 3189 + }, + { + "epoch": 2.547923322683706, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 3190 + }, + { + "epoch": 2.5487220447284344, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 3191 + }, + { + "epoch": 2.549520766773163, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 3192 + }, + { + "epoch": 2.5503194888178915, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 3193 + }, + { + "epoch": 2.5511182108626196, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 3194 + }, + { + "epoch": 2.551916932907348, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 3195 + }, + { + "epoch": 2.5527156549520766, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 3196 + }, + { + "epoch": 2.553514376996805, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1335, + "step": 3197 + }, + { + "epoch": 2.5543130990415337, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 3198 + }, + { + "epoch": 2.5551118210862622, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 3199 + }, + { + "epoch": 2.5559105431309903, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1295, + "step": 3200 + }, + { + "epoch": 2.556709265175719, + "grad_norm": 0.421875, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 3201 + }, + { + "epoch": 2.5575079872204474, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 3202 + }, + { + "epoch": 2.5583067092651754, + "grad_norm": 0.375, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 3203 + }, + { + "epoch": 2.559105431309904, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 3204 + }, + { + "epoch": 2.5599041533546325, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 3205 + }, + { + "epoch": 2.560702875399361, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 3206 + }, + { + "epoch": 2.5615015974440896, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1334, + "step": 3207 + }, + { + "epoch": 2.562300319488818, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 3208 + }, + { + "epoch": 2.563099041533546, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 3209 + }, + { + "epoch": 2.5638977635782747, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1401, + "step": 3210 + }, + { + "epoch": 2.5646964856230032, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 3211 + }, + { + "epoch": 2.5654952076677318, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 3212 + }, + { + "epoch": 2.56629392971246, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 3213 + }, + { + "epoch": 2.5670926517571884, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 3214 + }, + { + "epoch": 2.567891373801917, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 3215 + }, + { + "epoch": 2.5686900958466454, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 3216 + }, + { + "epoch": 2.569488817891374, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1371, + "step": 3217 + }, + { + "epoch": 2.5702875399361025, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 3218 + }, + { + "epoch": 2.5710862619808306, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 3219 + }, + { + "epoch": 2.571884984025559, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 3220 + }, + { + "epoch": 2.5726837060702876, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3221 + }, + { + "epoch": 2.5734824281150157, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 3222 + }, + { + "epoch": 2.5742811501597442, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 3223 + }, + { + "epoch": 2.5750798722044728, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 3224 + }, + { + "epoch": 2.5758785942492013, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 3225 + }, + { + "epoch": 2.57667731629393, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 3226 + }, + { + "epoch": 2.5774760383386583, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 3227 + }, + { + "epoch": 2.5782747603833864, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 3228 + }, + { + "epoch": 2.579073482428115, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 3229 + }, + { + "epoch": 2.5798722044728435, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 3230 + }, + { + "epoch": 2.580670926517572, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 3231 + }, + { + "epoch": 2.5814696485623, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 3232 + }, + { + "epoch": 2.5822683706070286, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 3233 + }, + { + "epoch": 2.583067092651757, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 3234 + }, + { + "epoch": 2.5838658146964857, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 3235 + }, + { + "epoch": 2.584664536741214, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 3236 + }, + { + "epoch": 2.5854632587859427, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.127, + "step": 3237 + }, + { + "epoch": 2.586261980830671, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3238 + }, + { + "epoch": 2.5870607028753994, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 3239 + }, + { + "epoch": 2.587859424920128, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 3240 + }, + { + "epoch": 2.588658146964856, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 3241 + }, + { + "epoch": 2.5894568690095845, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1353, + "step": 3242 + }, + { + "epoch": 2.590255591054313, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 3243 + }, + { + "epoch": 2.5910543130990416, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1355, + "step": 3244 + }, + { + "epoch": 2.59185303514377, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 3245 + }, + { + "epoch": 2.5926517571884986, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 3246 + }, + { + "epoch": 2.5934504792332267, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3247 + }, + { + "epoch": 2.594249201277955, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1273, + "step": 3248 + }, + { + "epoch": 2.5950479233226837, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 3249 + }, + { + "epoch": 2.5958466453674123, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 3250 + }, + { + "epoch": 2.5966453674121404, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1319, + "step": 3251 + }, + { + "epoch": 2.597444089456869, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 3252 + }, + { + "epoch": 2.5982428115015974, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 3253 + }, + { + "epoch": 2.599041533546326, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 3254 + }, + { + "epoch": 2.5998402555910545, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 3255 + }, + { + "epoch": 2.600638977635783, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 3256 + }, + { + "epoch": 2.601437699680511, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 3257 + }, + { + "epoch": 2.6022364217252396, + "grad_norm": 0.34375, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 3258 + }, + { + "epoch": 2.603035143769968, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 3259 + }, + { + "epoch": 2.6038338658146962, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 3260 + }, + { + "epoch": 2.6046325878594248, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 3261 + }, + { + "epoch": 2.6054313099041533, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3262 + }, + { + "epoch": 2.606230031948882, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 3263 + }, + { + "epoch": 2.6070287539936103, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3264 + }, + { + "epoch": 2.607827476038339, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 3265 + }, + { + "epoch": 2.608626198083067, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 3266 + }, + { + "epoch": 2.6094249201277955, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 3267 + }, + { + "epoch": 2.610223642172524, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 3268 + }, + { + "epoch": 2.6110223642172525, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 3269 + }, + { + "epoch": 2.6118210862619806, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 3270 + }, + { + "epoch": 2.612619808306709, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 3271 + }, + { + "epoch": 2.6134185303514377, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 3272 + }, + { + "epoch": 2.614217252396166, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 3273 + }, + { + "epoch": 2.6150159744408947, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1328, + "step": 3274 + }, + { + "epoch": 2.6158146964856233, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 3275 + }, + { + "epoch": 2.6166134185303513, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 3276 + }, + { + "epoch": 2.61741214057508, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 3277 + }, + { + "epoch": 2.6182108626198084, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 3278 + }, + { + "epoch": 2.6190095846645365, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 3279 + }, + { + "epoch": 2.619808306709265, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 3280 + }, + { + "epoch": 2.6206070287539935, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 3281 + }, + { + "epoch": 2.621405750798722, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 3282 + }, + { + "epoch": 2.6222044728434506, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 3283 + }, + { + "epoch": 2.623003194888179, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 3284 + }, + { + "epoch": 2.623801916932907, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 3285 + }, + { + "epoch": 2.6246006389776357, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 3286 + }, + { + "epoch": 2.6253993610223643, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 3287 + }, + { + "epoch": 2.626198083067093, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 3288 + }, + { + "epoch": 2.626996805111821, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 3289 + }, + { + "epoch": 2.6277955271565494, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 3290 + }, + { + "epoch": 2.628594249201278, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 3291 + }, + { + "epoch": 2.6293929712460065, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 3292 + }, + { + "epoch": 2.630191693290735, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 3293 + }, + { + "epoch": 2.6309904153354635, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1305, + "step": 3294 + }, + { + "epoch": 2.6317891373801916, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 3295 + }, + { + "epoch": 2.63258785942492, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 3296 + }, + { + "epoch": 2.6333865814696487, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3297 + }, + { + "epoch": 2.6341853035143767, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 3298 + }, + { + "epoch": 2.6349840255591053, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 3299 + }, + { + "epoch": 2.635782747603834, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3300 + }, + { + "epoch": 2.6365814696485623, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 3301 + }, + { + "epoch": 2.637380191693291, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 3302 + }, + { + "epoch": 2.6381789137380194, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 3303 + }, + { + "epoch": 2.6389776357827475, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 3304 + }, + { + "epoch": 2.639776357827476, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 3305 + }, + { + "epoch": 2.6405750798722045, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 3306 + }, + { + "epoch": 2.641373801916933, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1269, + "step": 3307 + }, + { + "epoch": 2.642172523961661, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 3308 + }, + { + "epoch": 2.6429712460063897, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 3309 + }, + { + "epoch": 2.643769968051118, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 3310 + }, + { + "epoch": 2.6445686900958467, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 3311 + }, + { + "epoch": 2.6453674121405752, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 3312 + }, + { + "epoch": 2.6461661341853038, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 3313 + }, + { + "epoch": 2.646964856230032, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 3314 + }, + { + "epoch": 2.6477635782747604, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 3315 + }, + { + "epoch": 2.648562300319489, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 3316 + }, + { + "epoch": 2.649361022364217, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 3317 + }, + { + "epoch": 2.6501597444089455, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 3318 + }, + { + "epoch": 2.650958466453674, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 3319 + }, + { + "epoch": 2.6517571884984026, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 3320 + }, + { + "epoch": 2.652555910543131, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 3321 + }, + { + "epoch": 2.6533546325878596, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 3322 + }, + { + "epoch": 2.6541533546325877, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 3323 + }, + { + "epoch": 2.6549520766773163, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3324 + }, + { + "epoch": 2.655750798722045, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 3325 + }, + { + "epoch": 2.6565495207667733, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 3326 + }, + { + "epoch": 2.6573482428115014, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 3327 + }, + { + "epoch": 2.65814696485623, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 3328 + }, + { + "epoch": 2.6589456869009584, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 3329 + }, + { + "epoch": 2.659744408945687, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 3330 + }, + { + "epoch": 2.6605431309904155, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 3331 + }, + { + "epoch": 2.661341853035144, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 3332 + }, + { + "epoch": 2.662140575079872, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 3333 + }, + { + "epoch": 2.6629392971246006, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1335, + "step": 3334 + }, + { + "epoch": 2.663738019169329, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 3335 + }, + { + "epoch": 2.6645367412140573, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3336 + }, + { + "epoch": 2.665335463258786, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3337 + }, + { + "epoch": 2.6661341853035143, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 3338 + }, + { + "epoch": 2.666932907348243, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1335, + "step": 3339 + }, + { + "epoch": 2.6677316293929714, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 3340 + }, + { + "epoch": 2.6685303514377, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 3341 + }, + { + "epoch": 2.669329073482428, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 3342 + }, + { + "epoch": 2.6701277955271565, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 3343 + }, + { + "epoch": 2.670926517571885, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 3344 + }, + { + "epoch": 2.6717252396166136, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 3345 + }, + { + "epoch": 2.6725239616613417, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 3346 + }, + { + "epoch": 2.67332268370607, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 3347 + }, + { + "epoch": 2.6741214057507987, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1305, + "step": 3348 + }, + { + "epoch": 2.6749201277955272, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 3349 + }, + { + "epoch": 2.6757188498402558, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 3350 + }, + { + "epoch": 2.6765175718849843, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 3351 + }, + { + "epoch": 2.6773162939297124, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1333, + "step": 3352 + }, + { + "epoch": 2.678115015974441, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 3353 + }, + { + "epoch": 2.6789137380191694, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1292, + "step": 3354 + }, + { + "epoch": 2.6797124600638975, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3355 + }, + { + "epoch": 2.680511182108626, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 3356 + }, + { + "epoch": 2.6813099041533546, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 3357 + }, + { + "epoch": 2.682108626198083, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 3358 + }, + { + "epoch": 2.6829073482428116, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 3359 + }, + { + "epoch": 2.68370607028754, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3360 + }, + { + "epoch": 2.6845047923322682, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1424, + "step": 3361 + }, + { + "epoch": 2.6853035143769968, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 3362 + }, + { + "epoch": 2.6861022364217253, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 3363 + }, + { + "epoch": 2.686900958466454, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 3364 + }, + { + "epoch": 2.687699680511182, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 3365 + }, + { + "epoch": 2.6884984025559104, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 3366 + }, + { + "epoch": 2.689297124600639, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 3367 + }, + { + "epoch": 2.6900958466453675, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 3368 + }, + { + "epoch": 2.690894568690096, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1311, + "step": 3369 + }, + { + "epoch": 2.6916932907348246, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 3370 + }, + { + "epoch": 2.6924920127795526, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 3371 + }, + { + "epoch": 2.693290734824281, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 3372 + }, + { + "epoch": 2.6940894568690097, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 3373 + }, + { + "epoch": 2.6948881789137378, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 3374 + }, + { + "epoch": 2.6956869009584663, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 3375 + }, + { + "epoch": 2.696485623003195, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 3376 + }, + { + "epoch": 2.6972843450479234, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 3377 + }, + { + "epoch": 2.698083067092652, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 3378 + }, + { + "epoch": 2.6988817891373804, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 3379 + }, + { + "epoch": 2.6996805111821085, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 3380 + }, + { + "epoch": 2.700479233226837, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.14, + "step": 3381 + }, + { + "epoch": 2.7012779552715656, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.128, + "step": 3382 + }, + { + "epoch": 2.702076677316294, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 3383 + }, + { + "epoch": 2.702875399361022, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 3384 + }, + { + "epoch": 2.7036741214057507, + "grad_norm": 0.39453125, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 3385 + }, + { + "epoch": 2.7044728434504792, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 3386 + }, + { + "epoch": 2.7052715654952078, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 3387 + }, + { + "epoch": 2.7060702875399363, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 3388 + }, + { + "epoch": 2.706869009584665, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1295, + "step": 3389 + }, + { + "epoch": 2.707667731629393, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1334, + "step": 3390 + }, + { + "epoch": 2.7084664536741214, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 3391 + }, + { + "epoch": 2.70926517571885, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 3392 + }, + { + "epoch": 2.710063897763578, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 3393 + }, + { + "epoch": 2.7108626198083066, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 3394 + }, + { + "epoch": 2.711661341853035, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 3395 + }, + { + "epoch": 2.7124600638977636, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 3396 + }, + { + "epoch": 2.713258785942492, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 3397 + }, + { + "epoch": 2.7140575079872207, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 3398 + }, + { + "epoch": 2.7148562300319488, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1328, + "step": 3399 + }, + { + "epoch": 2.7156549520766773, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1345, + "step": 3400 + }, + { + "epoch": 2.716453674121406, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 3401 + }, + { + "epoch": 2.7172523961661343, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 3402 + }, + { + "epoch": 2.7180511182108624, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1319, + "step": 3403 + }, + { + "epoch": 2.718849840255591, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 3404 + }, + { + "epoch": 2.7196485623003195, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 3405 + }, + { + "epoch": 2.720447284345048, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 3406 + }, + { + "epoch": 2.7212460063897765, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 3407 + }, + { + "epoch": 2.722044728434505, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1348, + "step": 3408 + }, + { + "epoch": 2.722843450479233, + "grad_norm": 0.37890625, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 3409 + }, + { + "epoch": 2.7236421725239617, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 3410 + }, + { + "epoch": 2.72444089456869, + "grad_norm": 0.388671875, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 3411 + }, + { + "epoch": 2.7252396166134183, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 3412 + }, + { + "epoch": 2.726038338658147, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1315, + "step": 3413 + }, + { + "epoch": 2.7268370607028753, + "grad_norm": 0.4765625, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 3414 + }, + { + "epoch": 2.727635782747604, + "grad_norm": 0.359375, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 3415 + }, + { + "epoch": 2.7284345047923324, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 3416 + }, + { + "epoch": 2.729233226837061, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 3417 + }, + { + "epoch": 2.730031948881789, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 3418 + }, + { + "epoch": 2.7308306709265175, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 3419 + }, + { + "epoch": 2.731629392971246, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1248, + "step": 3420 + }, + { + "epoch": 2.7324281150159746, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.131, + "step": 3421 + }, + { + "epoch": 2.7332268370607027, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.135, + "step": 3422 + }, + { + "epoch": 2.734025559105431, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.1283, + "step": 3423 + }, + { + "epoch": 2.7348242811501597, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 3424 + }, + { + "epoch": 2.7356230031948883, + "grad_norm": 0.34765625, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 3425 + }, + { + "epoch": 2.736421725239617, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 3426 + }, + { + "epoch": 2.737220447284345, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 3427 + }, + { + "epoch": 2.7380191693290734, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 3428 + }, + { + "epoch": 2.738817891373802, + "grad_norm": 0.34375, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 3429 + }, + { + "epoch": 2.7396166134185305, + "grad_norm": 0.48046875, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 3430 + }, + { + "epoch": 2.7404153354632586, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 3431 + }, + { + "epoch": 2.741214057507987, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 3432 + }, + { + "epoch": 2.7420127795527156, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 3433 + }, + { + "epoch": 2.742811501597444, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 3434 + }, + { + "epoch": 2.7436102236421727, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 3435 + }, + { + "epoch": 2.744408945686901, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3436 + }, + { + "epoch": 2.7452076677316293, + "grad_norm": 0.353515625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 3437 + }, + { + "epoch": 2.746006389776358, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3438 + }, + { + "epoch": 2.7468051118210863, + "grad_norm": 0.36328125, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 3439 + }, + { + "epoch": 2.747603833865815, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.127, + "step": 3440 + }, + { + "epoch": 2.748402555910543, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 3441 + }, + { + "epoch": 2.7492012779552715, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 3442 + }, + { + "epoch": 2.75, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 3443 + }, + { + "epoch": 2.7507987220447285, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 3444 + }, + { + "epoch": 2.751597444089457, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 3445 + }, + { + "epoch": 2.752396166134185, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1283, + "step": 3446 + }, + { + "epoch": 2.7531948881789137, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 3447 + }, + { + "epoch": 2.753993610223642, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 3448 + }, + { + "epoch": 2.7547923322683707, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3449 + }, + { + "epoch": 2.755591054313099, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 3450 + }, + { + "epoch": 2.7563897763578273, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 3451 + }, + { + "epoch": 2.757188498402556, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 3452 + }, + { + "epoch": 2.7579872204472844, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 3453 + }, + { + "epoch": 2.758785942492013, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 3454 + }, + { + "epoch": 2.7595846645367414, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3455 + }, + { + "epoch": 2.7603833865814695, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 3456 + }, + { + "epoch": 2.761182108626198, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3457 + }, + { + "epoch": 2.7619808306709266, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 3458 + }, + { + "epoch": 2.762779552715655, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 3459 + }, + { + "epoch": 2.763578274760383, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 3460 + }, + { + "epoch": 2.7643769968051117, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1273, + "step": 3461 + }, + { + "epoch": 2.7651757188498403, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 3462 + }, + { + "epoch": 2.765974440894569, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 3463 + }, + { + "epoch": 2.7667731629392973, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 3464 + }, + { + "epoch": 2.7675718849840254, + "grad_norm": 0.419921875, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3465 + }, + { + "epoch": 2.768370607028754, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 3466 + }, + { + "epoch": 2.7691693290734825, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 3467 + }, + { + "epoch": 2.769968051118211, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 3468 + }, + { + "epoch": 2.770766773162939, + "grad_norm": 0.41015625, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 3469 + }, + { + "epoch": 2.7715654952076676, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 3470 + }, + { + "epoch": 2.772364217252396, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 3471 + }, + { + "epoch": 2.7731629392971247, + "grad_norm": 0.455078125, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 3472 + }, + { + "epoch": 2.773961661341853, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1269, + "step": 3473 + }, + { + "epoch": 2.7747603833865817, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 3474 + }, + { + "epoch": 2.77555910543131, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3475 + }, + { + "epoch": 2.7763578274760383, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 3476 + }, + { + "epoch": 2.777156549520767, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 3477 + }, + { + "epoch": 2.777955271565495, + "grad_norm": 0.40234375, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 3478 + }, + { + "epoch": 2.7787539936102235, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 3479 + }, + { + "epoch": 2.779552715654952, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 3480 + }, + { + "epoch": 2.7803514376996805, + "grad_norm": 0.451171875, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 3481 + }, + { + "epoch": 2.781150159744409, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 3482 + }, + { + "epoch": 2.7819488817891376, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1358, + "step": 3483 + }, + { + "epoch": 2.7827476038338657, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 3484 + }, + { + "epoch": 2.783546325878594, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3485 + }, + { + "epoch": 2.7843450479233227, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 3486 + }, + { + "epoch": 2.7851437699680512, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3487 + }, + { + "epoch": 2.7859424920127793, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 3488 + }, + { + "epoch": 2.786741214057508, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 3489 + }, + { + "epoch": 2.7875399361022364, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 3490 + }, + { + "epoch": 2.788338658146965, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 3491 + }, + { + "epoch": 2.7891373801916934, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 3492 + }, + { + "epoch": 2.789936102236422, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 3493 + }, + { + "epoch": 2.79073482428115, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 3494 + }, + { + "epoch": 2.7915335463258786, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 3495 + }, + { + "epoch": 2.792332268370607, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 3496 + }, + { + "epoch": 2.793130990415335, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 3497 + }, + { + "epoch": 2.7939297124600637, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 3498 + }, + { + "epoch": 2.7947284345047922, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 3499 + }, + { + "epoch": 2.7955271565495208, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 3500 + }, + { + "epoch": 2.7963258785942493, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 3501 + }, + { + "epoch": 2.797124600638978, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 3502 + }, + { + "epoch": 2.797923322683706, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 3503 + }, + { + "epoch": 2.7987220447284344, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 3504 + }, + { + "epoch": 2.799520766773163, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 3505 + }, + { + "epoch": 2.8003194888178915, + "grad_norm": 0.490234375, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 3506 + }, + { + "epoch": 2.8011182108626196, + "grad_norm": 0.353515625, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 3507 + }, + { + "epoch": 2.801916932907348, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 3508 + }, + { + "epoch": 2.8027156549520766, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 3509 + }, + { + "epoch": 2.803514376996805, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3510 + }, + { + "epoch": 2.8043130990415337, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 3511 + }, + { + "epoch": 2.8051118210862622, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.127, + "step": 3512 + }, + { + "epoch": 2.8059105431309903, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 3513 + }, + { + "epoch": 2.806709265175719, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 3514 + }, + { + "epoch": 2.8075079872204474, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 3515 + }, + { + "epoch": 2.8083067092651754, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 3516 + }, + { + "epoch": 2.809105431309904, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 3517 + }, + { + "epoch": 2.8099041533546325, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 3518 + }, + { + "epoch": 2.810702875399361, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 3519 + }, + { + "epoch": 2.8115015974440896, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 3520 + }, + { + "epoch": 2.812300319488818, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 3521 + }, + { + "epoch": 2.813099041533546, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 3522 + }, + { + "epoch": 2.8138977635782747, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 3523 + }, + { + "epoch": 2.8146964856230032, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 3524 + }, + { + "epoch": 2.8154952076677318, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3525 + }, + { + "epoch": 2.81629392971246, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1365, + "step": 3526 + }, + { + "epoch": 2.8170926517571884, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 3527 + }, + { + "epoch": 2.817891373801917, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 3528 + }, + { + "epoch": 2.8186900958466454, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 3529 + }, + { + "epoch": 2.819488817891374, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 3530 + }, + { + "epoch": 2.8202875399361025, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 3531 + }, + { + "epoch": 2.8210862619808306, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 3532 + }, + { + "epoch": 2.821884984025559, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 3533 + }, + { + "epoch": 2.8226837060702876, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 3534 + }, + { + "epoch": 2.8234824281150157, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 3535 + }, + { + "epoch": 2.8242811501597442, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 3536 + }, + { + "epoch": 2.8250798722044728, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 3537 + }, + { + "epoch": 2.8258785942492013, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 3538 + }, + { + "epoch": 2.82667731629393, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 3539 + }, + { + "epoch": 2.8274760383386583, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3540 + }, + { + "epoch": 2.8282747603833864, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 3541 + }, + { + "epoch": 2.829073482428115, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 3542 + }, + { + "epoch": 2.8298722044728435, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.133, + "step": 3543 + }, + { + "epoch": 2.830670926517572, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 3544 + }, + { + "epoch": 2.8314696485623, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3545 + }, + { + "epoch": 2.8322683706070286, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 3546 + }, + { + "epoch": 2.833067092651757, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 3547 + }, + { + "epoch": 2.8338658146964857, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 3548 + }, + { + "epoch": 2.834664536741214, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 3549 + }, + { + "epoch": 2.8354632587859427, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 3550 + }, + { + "epoch": 2.836261980830671, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 3551 + }, + { + "epoch": 2.8370607028753994, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 3552 + }, + { + "epoch": 2.837859424920128, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 3553 + }, + { + "epoch": 2.838658146964856, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 3554 + }, + { + "epoch": 2.8394568690095845, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3555 + }, + { + "epoch": 2.840255591054313, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1273, + "step": 3556 + }, + { + "epoch": 2.8410543130990416, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 3557 + }, + { + "epoch": 2.84185303514377, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3558 + }, + { + "epoch": 2.8426517571884986, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 3559 + }, + { + "epoch": 2.8434504792332267, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 3560 + }, + { + "epoch": 2.844249201277955, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 3561 + }, + { + "epoch": 2.8450479233226837, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 3562 + }, + { + "epoch": 2.8458466453674123, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1319, + "step": 3563 + }, + { + "epoch": 2.8466453674121404, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 3564 + }, + { + "epoch": 2.847444089456869, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 3565 + }, + { + "epoch": 2.8482428115015974, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 3566 + }, + { + "epoch": 2.849041533546326, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 3567 + }, + { + "epoch": 2.8498402555910545, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3568 + }, + { + "epoch": 2.850638977635783, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 3569 + }, + { + "epoch": 2.851437699680511, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 3570 + }, + { + "epoch": 2.8522364217252396, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 3571 + }, + { + "epoch": 2.853035143769968, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3572 + }, + { + "epoch": 2.8538338658146962, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 3573 + }, + { + "epoch": 2.8546325878594248, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 3574 + }, + { + "epoch": 2.8554313099041533, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 3575 + }, + { + "epoch": 2.856230031948882, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 3576 + }, + { + "epoch": 2.8570287539936103, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 3577 + }, + { + "epoch": 2.857827476038339, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1313, + "step": 3578 + }, + { + "epoch": 2.858626198083067, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 3579 + }, + { + "epoch": 2.8594249201277955, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 3580 + }, + { + "epoch": 2.860223642172524, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 3581 + }, + { + "epoch": 2.8610223642172525, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 3582 + }, + { + "epoch": 2.8618210862619806, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 3583 + }, + { + "epoch": 2.862619808306709, + "grad_norm": 0.380859375, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3584 + }, + { + "epoch": 2.8634185303514377, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 3585 + }, + { + "epoch": 2.864217252396166, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 3586 + }, + { + "epoch": 2.8650159744408947, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 3587 + }, + { + "epoch": 2.8658146964856233, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 3588 + }, + { + "epoch": 2.8666134185303513, + "grad_norm": 0.380859375, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 3589 + }, + { + "epoch": 2.86741214057508, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1339, + "step": 3590 + }, + { + "epoch": 2.8682108626198084, + "grad_norm": 0.34765625, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 3591 + }, + { + "epoch": 2.8690095846645365, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3592 + }, + { + "epoch": 2.869808306709265, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 3593 + }, + { + "epoch": 2.8706070287539935, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 3594 + }, + { + "epoch": 2.871405750798722, + "grad_norm": 0.396484375, + "learning_rate": 0.0001, + "loss": 1.1355, + "step": 3595 + }, + { + "epoch": 2.8722044728434506, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 3596 + }, + { + "epoch": 2.873003194888179, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 3597 + }, + { + "epoch": 2.873801916932907, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 3598 + }, + { + "epoch": 2.8746006389776357, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 3599 + }, + { + "epoch": 2.8753993610223643, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 3600 + }, + { + "epoch": 2.876198083067093, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 3601 + }, + { + "epoch": 2.876996805111821, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 3602 + }, + { + "epoch": 2.8777955271565494, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 3603 + }, + { + "epoch": 2.878594249201278, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 3604 + }, + { + "epoch": 2.8793929712460065, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 3605 + }, + { + "epoch": 2.880191693290735, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 3606 + }, + { + "epoch": 2.8809904153354635, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 3607 + }, + { + "epoch": 2.8817891373801916, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1269, + "step": 3608 + }, + { + "epoch": 2.88258785942492, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 3609 + }, + { + "epoch": 2.8833865814696487, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 3610 + }, + { + "epoch": 2.8841853035143767, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 3611 + }, + { + "epoch": 2.8849840255591053, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 3612 + }, + { + "epoch": 2.885782747603834, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 3613 + }, + { + "epoch": 2.8865814696485623, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 3614 + }, + { + "epoch": 2.887380191693291, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 3615 + }, + { + "epoch": 2.8881789137380194, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1294, + "step": 3616 + }, + { + "epoch": 2.8889776357827475, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 3617 + }, + { + "epoch": 2.889776357827476, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 3618 + }, + { + "epoch": 2.8905750798722045, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 3619 + }, + { + "epoch": 2.891373801916933, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1343, + "step": 3620 + }, + { + "epoch": 2.892172523961661, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 3621 + }, + { + "epoch": 2.8929712460063897, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 3622 + }, + { + "epoch": 2.893769968051118, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3623 + }, + { + "epoch": 2.8945686900958467, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1337, + "step": 3624 + }, + { + "epoch": 2.8953674121405752, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 3625 + }, + { + "epoch": 2.8961661341853038, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 3626 + }, + { + "epoch": 2.896964856230032, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 3627 + }, + { + "epoch": 2.8977635782747604, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.128, + "step": 3628 + }, + { + "epoch": 2.898562300319489, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 3629 + }, + { + "epoch": 2.899361022364217, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3630 + }, + { + "epoch": 2.9001597444089455, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 3631 + }, + { + "epoch": 2.900958466453674, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 3632 + }, + { + "epoch": 2.9017571884984026, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 3633 + }, + { + "epoch": 2.902555910543131, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 3634 + }, + { + "epoch": 2.9033546325878596, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 3635 + }, + { + "epoch": 2.9041533546325877, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1235, + "step": 3636 + }, + { + "epoch": 2.9049520766773163, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 3637 + }, + { + "epoch": 2.905750798722045, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 3638 + }, + { + "epoch": 2.9065495207667733, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 3639 + }, + { + "epoch": 2.9073482428115014, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 3640 + }, + { + "epoch": 2.90814696485623, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 3641 + }, + { + "epoch": 2.9089456869009584, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 3642 + }, + { + "epoch": 2.909744408945687, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 3643 + }, + { + "epoch": 2.9105431309904155, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1302, + "step": 3644 + }, + { + "epoch": 2.911341853035144, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 3645 + }, + { + "epoch": 2.912140575079872, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 3646 + }, + { + "epoch": 2.9129392971246006, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 3647 + }, + { + "epoch": 2.913738019169329, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 3648 + }, + { + "epoch": 2.9145367412140573, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 3649 + }, + { + "epoch": 2.915335463258786, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 3650 + }, + { + "epoch": 2.9161341853035143, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1287, + "step": 3651 + }, + { + "epoch": 2.916932907348243, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 3652 + }, + { + "epoch": 2.9177316293929714, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3653 + }, + { + "epoch": 2.9185303514377, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 3654 + }, + { + "epoch": 2.919329073482428, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 3655 + }, + { + "epoch": 2.9201277955271565, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1399, + "step": 3656 + }, + { + "epoch": 2.920926517571885, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 3657 + }, + { + "epoch": 2.9217252396166136, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 3658 + }, + { + "epoch": 2.9225239616613417, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 3659 + }, + { + "epoch": 2.92332268370607, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 3660 + }, + { + "epoch": 2.9241214057507987, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 3661 + }, + { + "epoch": 2.9249201277955272, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 3662 + }, + { + "epoch": 2.9257188498402558, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 3663 + }, + { + "epoch": 2.9265175718849843, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 3664 + }, + { + "epoch": 2.9273162939297124, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 3665 + }, + { + "epoch": 2.928115015974441, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 3666 + }, + { + "epoch": 2.9289137380191694, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 3667 + }, + { + "epoch": 2.9297124600638975, + "grad_norm": 0.33984375, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 3668 + }, + { + "epoch": 2.930511182108626, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 3669 + }, + { + "epoch": 2.9313099041533546, + "grad_norm": 0.54296875, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 3670 + }, + { + "epoch": 2.932108626198083, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 3671 + }, + { + "epoch": 2.9329073482428116, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 3672 + }, + { + "epoch": 2.93370607028754, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 3673 + }, + { + "epoch": 2.9345047923322682, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 3674 + }, + { + "epoch": 2.9353035143769968, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 3675 + }, + { + "epoch": 2.9361022364217253, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3676 + }, + { + "epoch": 2.936900958466454, + "grad_norm": 0.37890625, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3677 + }, + { + "epoch": 2.937699680511182, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 3678 + }, + { + "epoch": 2.9384984025559104, + "grad_norm": 0.369140625, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 3679 + }, + { + "epoch": 2.939297124600639, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 3680 + }, + { + "epoch": 2.9400958466453675, + "grad_norm": 0.3828125, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 3681 + }, + { + "epoch": 2.940894568690096, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1305, + "step": 3682 + }, + { + "epoch": 2.9416932907348246, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 3683 + }, + { + "epoch": 2.9424920127795526, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 3684 + }, + { + "epoch": 2.943290734824281, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 3685 + }, + { + "epoch": 2.9440894568690097, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 3686 + }, + { + "epoch": 2.9448881789137378, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 3687 + }, + { + "epoch": 2.9456869009584663, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 3688 + }, + { + "epoch": 2.946485623003195, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 3689 + }, + { + "epoch": 2.9472843450479234, + "grad_norm": 0.318359375, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 3690 + }, + { + "epoch": 2.948083067092652, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 3691 + }, + { + "epoch": 2.9488817891373804, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 3692 + }, + { + "epoch": 2.9496805111821085, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 3693 + }, + { + "epoch": 2.950479233226837, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 3694 + }, + { + "epoch": 2.9512779552715656, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 3695 + }, + { + "epoch": 2.952076677316294, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 3696 + }, + { + "epoch": 2.952875399361022, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 3697 + }, + { + "epoch": 2.9536741214057507, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 3698 + }, + { + "epoch": 2.9544728434504792, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 3699 + }, + { + "epoch": 2.9552715654952078, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 3700 + }, + { + "epoch": 2.9560702875399363, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 3701 + }, + { + "epoch": 2.956869009584665, + "grad_norm": 0.35546875, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 3702 + }, + { + "epoch": 2.957667731629393, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 3703 + }, + { + "epoch": 2.9584664536741214, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 3704 + }, + { + "epoch": 2.95926517571885, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 3705 + }, + { + "epoch": 2.960063897763578, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 3706 + }, + { + "epoch": 2.9608626198083066, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1334, + "step": 3707 + }, + { + "epoch": 2.961661341853035, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3708 + }, + { + "epoch": 2.9624600638977636, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 3709 + }, + { + "epoch": 2.963258785942492, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1334, + "step": 3710 + }, + { + "epoch": 2.9640575079872207, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 3711 + }, + { + "epoch": 2.9648562300319488, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 3712 + }, + { + "epoch": 2.9656549520766773, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 3713 + }, + { + "epoch": 2.966453674121406, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 3714 + }, + { + "epoch": 2.9672523961661343, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 3715 + }, + { + "epoch": 2.9680511182108624, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 3716 + }, + { + "epoch": 2.968849840255591, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1317, + "step": 3717 + }, + { + "epoch": 2.9696485623003195, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 3718 + }, + { + "epoch": 2.970447284345048, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1287, + "step": 3719 + }, + { + "epoch": 2.9712460063897765, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 3720 + }, + { + "epoch": 2.972044728434505, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 3721 + }, + { + "epoch": 2.972843450479233, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3722 + }, + { + "epoch": 2.9736421725239617, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 3723 + }, + { + "epoch": 2.97444089456869, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 3724 + }, + { + "epoch": 2.9752396166134183, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 3725 + }, + { + "epoch": 2.976038338658147, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 3726 + }, + { + "epoch": 2.9768370607028753, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 3727 + }, + { + "epoch": 2.977635782747604, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 3728 + }, + { + "epoch": 2.9784345047923324, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 3729 + }, + { + "epoch": 2.979233226837061, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 3730 + }, + { + "epoch": 2.980031948881789, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.134, + "step": 3731 + }, + { + "epoch": 2.9808306709265175, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 3732 + }, + { + "epoch": 2.981629392971246, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 3733 + }, + { + "epoch": 2.9824281150159746, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3734 + }, + { + "epoch": 2.9832268370607027, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 3735 + }, + { + "epoch": 2.984025559105431, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 3736 + }, + { + "epoch": 2.9848242811501597, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 3737 + }, + { + "epoch": 2.9856230031948883, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 3738 + }, + { + "epoch": 2.986421725239617, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 3739 + }, + { + "epoch": 2.987220447284345, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 3740 + }, + { + "epoch": 2.9880191693290734, + "grad_norm": 0.5, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 3741 + }, + { + "epoch": 2.988817891373802, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 3742 + }, + { + "epoch": 2.9896166134185305, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 3743 + }, + { + "epoch": 2.9904153354632586, + "grad_norm": 0.388671875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 3744 + }, + { + "epoch": 2.991214057507987, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 3745 + }, + { + "epoch": 2.9920127795527156, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1323, + "step": 3746 + }, + { + "epoch": 2.992811501597444, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 3747 + }, + { + "epoch": 2.9936102236421727, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 3748 + }, + { + "epoch": 2.994408945686901, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 3749 + }, + { + "epoch": 2.9952076677316293, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 3750 + }, + { + "epoch": 2.996006389776358, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 3751 + }, + { + "epoch": 2.9968051118210863, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3752 + }, + { + "epoch": 2.997603833865815, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 3753 + }, + { + "epoch": 2.998402555910543, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3754 + }, + { + "epoch": 2.9992012779552715, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 3755 + }, + { + "epoch": 3.0, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 3756 + }, + { + "epoch": 3.0007987220447285, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 3757 + }, + { + "epoch": 3.001597444089457, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 3758 + }, + { + "epoch": 3.002396166134185, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 3759 + }, + { + "epoch": 3.0031948881789137, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 3760 + }, + { + "epoch": 3.003993610223642, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 3761 + }, + { + "epoch": 3.0047923322683707, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1305, + "step": 3762 + }, + { + "epoch": 3.0055910543130993, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 3763 + }, + { + "epoch": 3.0063897763578273, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1342, + "step": 3764 + }, + { + "epoch": 3.007188498402556, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 3765 + }, + { + "epoch": 3.0079872204472844, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 3766 + }, + { + "epoch": 3.008785942492013, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 3767 + }, + { + "epoch": 3.009584664536741, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 3768 + }, + { + "epoch": 3.0103833865814695, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 3769 + }, + { + "epoch": 3.011182108626198, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 3770 + }, + { + "epoch": 3.0119808306709266, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 3771 + }, + { + "epoch": 3.012779552715655, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 3772 + }, + { + "epoch": 3.013578274760383, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 3773 + }, + { + "epoch": 3.0143769968051117, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 3774 + }, + { + "epoch": 3.0151757188498403, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3775 + }, + { + "epoch": 3.015974440894569, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 3776 + }, + { + "epoch": 3.0167731629392973, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 3777 + }, + { + "epoch": 3.0175718849840254, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 3778 + }, + { + "epoch": 3.018370607028754, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 3779 + }, + { + "epoch": 3.0191693290734825, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 3780 + }, + { + "epoch": 3.019968051118211, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 3781 + }, + { + "epoch": 3.0207667731629395, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 3782 + }, + { + "epoch": 3.0215654952076676, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 3783 + }, + { + "epoch": 3.022364217252396, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 3784 + }, + { + "epoch": 3.0231629392971247, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 3785 + }, + { + "epoch": 3.023961661341853, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 3786 + }, + { + "epoch": 3.0247603833865813, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 3787 + }, + { + "epoch": 3.02555910543131, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 3788 + }, + { + "epoch": 3.0263578274760383, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.128, + "step": 3789 + }, + { + "epoch": 3.027156549520767, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 3790 + }, + { + "epoch": 3.0279552715654954, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 3791 + }, + { + "epoch": 3.0287539936102235, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 3792 + }, + { + "epoch": 3.029552715654952, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 3793 + }, + { + "epoch": 3.0303514376996805, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 3794 + }, + { + "epoch": 3.031150159744409, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 3795 + }, + { + "epoch": 3.0319488817891376, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 3796 + }, + { + "epoch": 3.0327476038338657, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 3797 + }, + { + "epoch": 3.033546325878594, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 3798 + }, + { + "epoch": 3.0343450479233227, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 3799 + }, + { + "epoch": 3.0351437699680512, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 3800 + }, + { + "epoch": 3.0359424920127798, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3801 + }, + { + "epoch": 3.036741214057508, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 3802 + }, + { + "epoch": 3.0375399361022364, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 3803 + }, + { + "epoch": 3.038338658146965, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 3804 + }, + { + "epoch": 3.0391373801916934, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 3805 + }, + { + "epoch": 3.0399361022364215, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 3806 + }, + { + "epoch": 3.04073482428115, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 3807 + }, + { + "epoch": 3.0415335463258786, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 3808 + }, + { + "epoch": 3.042332268370607, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 3809 + }, + { + "epoch": 3.0431309904153356, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 3810 + }, + { + "epoch": 3.0439297124600637, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 3811 + }, + { + "epoch": 3.0447284345047922, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 3812 + }, + { + "epoch": 3.0455271565495208, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 3813 + }, + { + "epoch": 3.0463258785942493, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1291, + "step": 3814 + }, + { + "epoch": 3.047124600638978, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 3815 + }, + { + "epoch": 3.047923322683706, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3816 + }, + { + "epoch": 3.0487220447284344, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 3817 + }, + { + "epoch": 3.049520766773163, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 3818 + }, + { + "epoch": 3.0503194888178915, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 3819 + }, + { + "epoch": 3.0511182108626196, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 3820 + }, + { + "epoch": 3.051916932907348, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 3821 + }, + { + "epoch": 3.0527156549520766, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 3822 + }, + { + "epoch": 3.053514376996805, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1269, + "step": 3823 + }, + { + "epoch": 3.0543130990415337, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 3824 + }, + { + "epoch": 3.055111821086262, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 3825 + }, + { + "epoch": 3.0559105431309903, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 3826 + }, + { + "epoch": 3.056709265175719, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 3827 + }, + { + "epoch": 3.0575079872204474, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1347, + "step": 3828 + }, + { + "epoch": 3.058306709265176, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 3829 + }, + { + "epoch": 3.059105431309904, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 3830 + }, + { + "epoch": 3.0599041533546325, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 3831 + }, + { + "epoch": 3.060702875399361, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 3832 + }, + { + "epoch": 3.0615015974440896, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 3833 + }, + { + "epoch": 3.062300319488818, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 3834 + }, + { + "epoch": 3.063099041533546, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 3835 + }, + { + "epoch": 3.0638977635782747, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 3836 + }, + { + "epoch": 3.0646964856230032, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 3837 + }, + { + "epoch": 3.0654952076677318, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 3838 + }, + { + "epoch": 3.06629392971246, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 3839 + }, + { + "epoch": 3.0670926517571884, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 3840 + }, + { + "epoch": 3.067891373801917, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1308, + "step": 3841 + }, + { + "epoch": 3.0686900958466454, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 3842 + }, + { + "epoch": 3.069488817891374, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 3843 + }, + { + "epoch": 3.070287539936102, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 3844 + }, + { + "epoch": 3.0710862619808306, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 3845 + }, + { + "epoch": 3.071884984025559, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 3846 + }, + { + "epoch": 3.0726837060702876, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 3847 + }, + { + "epoch": 3.073482428115016, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 3848 + }, + { + "epoch": 3.0742811501597442, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 3849 + }, + { + "epoch": 3.0750798722044728, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 3850 + }, + { + "epoch": 3.0758785942492013, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1273, + "step": 3851 + }, + { + "epoch": 3.07667731629393, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 3852 + }, + { + "epoch": 3.0774760383386583, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 3853 + }, + { + "epoch": 3.0782747603833864, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 3854 + }, + { + "epoch": 3.079073482428115, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 3855 + }, + { + "epoch": 3.0798722044728435, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 3856 + }, + { + "epoch": 3.080670926517572, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 3857 + }, + { + "epoch": 3.0814696485623, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 3858 + }, + { + "epoch": 3.0822683706070286, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 3859 + }, + { + "epoch": 3.083067092651757, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 3860 + }, + { + "epoch": 3.0838658146964857, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 3861 + }, + { + "epoch": 3.084664536741214, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 3862 + }, + { + "epoch": 3.0854632587859423, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 3863 + }, + { + "epoch": 3.086261980830671, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.13, + "step": 3864 + }, + { + "epoch": 3.0870607028753994, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 3865 + }, + { + "epoch": 3.087859424920128, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 3866 + }, + { + "epoch": 3.0886581469648564, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1311, + "step": 3867 + }, + { + "epoch": 3.0894568690095845, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 3868 + }, + { + "epoch": 3.090255591054313, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 3869 + }, + { + "epoch": 3.0910543130990416, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3870 + }, + { + "epoch": 3.09185303514377, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 3871 + }, + { + "epoch": 3.0926517571884986, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 3872 + }, + { + "epoch": 3.0934504792332267, + "grad_norm": 0.59765625, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 3873 + }, + { + "epoch": 3.094249201277955, + "grad_norm": 0.41015625, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 3874 + }, + { + "epoch": 3.0950479233226837, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 3875 + }, + { + "epoch": 3.0958466453674123, + "grad_norm": 0.36328125, + "learning_rate": 0.0001, + "loss": 1.1328, + "step": 3876 + }, + { + "epoch": 3.0966453674121404, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 3877 + }, + { + "epoch": 3.097444089456869, + "grad_norm": 0.392578125, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3878 + }, + { + "epoch": 3.0982428115015974, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 3879 + }, + { + "epoch": 3.099041533546326, + "grad_norm": 0.37109375, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 3880 + }, + { + "epoch": 3.0998402555910545, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 3881 + }, + { + "epoch": 3.1006389776357826, + "grad_norm": 0.35546875, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 3882 + }, + { + "epoch": 3.101437699680511, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3883 + }, + { + "epoch": 3.1022364217252396, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 3884 + }, + { + "epoch": 3.103035143769968, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 3885 + }, + { + "epoch": 3.1038338658146967, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 3886 + }, + { + "epoch": 3.1046325878594248, + "grad_norm": 0.53515625, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 3887 + }, + { + "epoch": 3.1054313099041533, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 3888 + }, + { + "epoch": 3.106230031948882, + "grad_norm": 0.3828125, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3889 + }, + { + "epoch": 3.1070287539936103, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 3890 + }, + { + "epoch": 3.107827476038339, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 3891 + }, + { + "epoch": 3.108626198083067, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 3892 + }, + { + "epoch": 3.1094249201277955, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1407, + "step": 3893 + }, + { + "epoch": 3.110223642172524, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1303, + "step": 3894 + }, + { + "epoch": 3.1110223642172525, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3895 + }, + { + "epoch": 3.1118210862619806, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 3896 + }, + { + "epoch": 3.112619808306709, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 3897 + }, + { + "epoch": 3.1134185303514377, + "grad_norm": 0.373046875, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 3898 + }, + { + "epoch": 3.114217252396166, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 3899 + }, + { + "epoch": 3.1150159744408947, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 3900 + }, + { + "epoch": 3.115814696485623, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 3901 + }, + { + "epoch": 3.1166134185303513, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 3902 + }, + { + "epoch": 3.11741214057508, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 3903 + }, + { + "epoch": 3.1182108626198084, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 3904 + }, + { + "epoch": 3.119009584664537, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 3905 + }, + { + "epoch": 3.119808306709265, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 3906 + }, + { + "epoch": 3.1206070287539935, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 3907 + }, + { + "epoch": 3.121405750798722, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 3908 + }, + { + "epoch": 3.1222044728434506, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3909 + }, + { + "epoch": 3.123003194888179, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 3910 + }, + { + "epoch": 3.123801916932907, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 3911 + }, + { + "epoch": 3.1246006389776357, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 3912 + }, + { + "epoch": 3.1253993610223643, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 3913 + }, + { + "epoch": 3.126198083067093, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3914 + }, + { + "epoch": 3.126996805111821, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 3915 + }, + { + "epoch": 3.1277955271565494, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 3916 + }, + { + "epoch": 3.128594249201278, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 3917 + }, + { + "epoch": 3.1293929712460065, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 3918 + }, + { + "epoch": 3.130191693290735, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 3919 + }, + { + "epoch": 3.130990415335463, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 3920 + }, + { + "epoch": 3.1317891373801916, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 3921 + }, + { + "epoch": 3.13258785942492, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 3922 + }, + { + "epoch": 3.1333865814696487, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 3923 + }, + { + "epoch": 3.134185303514377, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 3924 + }, + { + "epoch": 3.1349840255591053, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 3925 + }, + { + "epoch": 3.135782747603834, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 3926 + }, + { + "epoch": 3.1365814696485623, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 3927 + }, + { + "epoch": 3.137380191693291, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 3928 + }, + { + "epoch": 3.1381789137380194, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 3929 + }, + { + "epoch": 3.1389776357827475, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 3930 + }, + { + "epoch": 3.139776357827476, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 3931 + }, + { + "epoch": 3.1405750798722045, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 3932 + }, + { + "epoch": 3.141373801916933, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 3933 + }, + { + "epoch": 3.142172523961661, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 3934 + }, + { + "epoch": 3.1429712460063897, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 3935 + }, + { + "epoch": 3.143769968051118, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 3936 + }, + { + "epoch": 3.1445686900958467, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1287, + "step": 3937 + }, + { + "epoch": 3.1453674121405752, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 3938 + }, + { + "epoch": 3.1461661341853033, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 3939 + }, + { + "epoch": 3.146964856230032, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 3940 + }, + { + "epoch": 3.1477635782747604, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3941 + }, + { + "epoch": 3.148562300319489, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 3942 + }, + { + "epoch": 3.1493610223642174, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 3943 + }, + { + "epoch": 3.1501597444089455, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 3944 + }, + { + "epoch": 3.150958466453674, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 3945 + }, + { + "epoch": 3.1517571884984026, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 3946 + }, + { + "epoch": 3.152555910543131, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 3947 + }, + { + "epoch": 3.1533546325878596, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 3948 + }, + { + "epoch": 3.1541533546325877, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 3949 + }, + { + "epoch": 3.1549520766773163, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 3950 + }, + { + "epoch": 3.155750798722045, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 3951 + }, + { + "epoch": 3.1565495207667733, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 3952 + }, + { + "epoch": 3.1573482428115014, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 3953 + }, + { + "epoch": 3.15814696485623, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 3954 + }, + { + "epoch": 3.1589456869009584, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 3955 + }, + { + "epoch": 3.159744408945687, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 3956 + }, + { + "epoch": 3.1605431309904155, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 3957 + }, + { + "epoch": 3.1613418530351436, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 3958 + }, + { + "epoch": 3.162140575079872, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 3959 + }, + { + "epoch": 3.1629392971246006, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 3960 + }, + { + "epoch": 3.163738019169329, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 3961 + }, + { + "epoch": 3.1645367412140577, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1273, + "step": 3962 + }, + { + "epoch": 3.165335463258786, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 3963 + }, + { + "epoch": 3.1661341853035143, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 3964 + }, + { + "epoch": 3.166932907348243, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 3965 + }, + { + "epoch": 3.1677316293929714, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 3966 + }, + { + "epoch": 3.1685303514377, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 3967 + }, + { + "epoch": 3.169329073482428, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 3968 + }, + { + "epoch": 3.1701277955271565, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1333, + "step": 3969 + }, + { + "epoch": 3.170926517571885, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3970 + }, + { + "epoch": 3.1717252396166136, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 3971 + }, + { + "epoch": 3.1725239616613417, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 3972 + }, + { + "epoch": 3.17332268370607, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1235, + "step": 3973 + }, + { + "epoch": 3.1741214057507987, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 3974 + }, + { + "epoch": 3.1749201277955272, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 3975 + }, + { + "epoch": 3.1757188498402558, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 3976 + }, + { + "epoch": 3.176517571884984, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 3977 + }, + { + "epoch": 3.1773162939297124, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 3978 + }, + { + "epoch": 3.178115015974441, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 3979 + }, + { + "epoch": 3.1789137380191694, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 3980 + }, + { + "epoch": 3.179712460063898, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 3981 + }, + { + "epoch": 3.180511182108626, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 3982 + }, + { + "epoch": 3.1813099041533546, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 3983 + }, + { + "epoch": 3.182108626198083, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 3984 + }, + { + "epoch": 3.1829073482428116, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 3985 + }, + { + "epoch": 3.18370607028754, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 3986 + }, + { + "epoch": 3.1845047923322682, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 3987 + }, + { + "epoch": 3.1853035143769968, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 3988 + }, + { + "epoch": 3.1861022364217253, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 3989 + }, + { + "epoch": 3.186900958466454, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 3990 + }, + { + "epoch": 3.187699680511182, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3991 + }, + { + "epoch": 3.1884984025559104, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 3992 + }, + { + "epoch": 3.189297124600639, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 3993 + }, + { + "epoch": 3.1900958466453675, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 3994 + }, + { + "epoch": 3.190894568690096, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 3995 + }, + { + "epoch": 3.191693290734824, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 3996 + }, + { + "epoch": 3.1924920127795526, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 3997 + }, + { + "epoch": 3.193290734824281, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 3998 + }, + { + "epoch": 3.1940894568690097, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 3999 + }, + { + "epoch": 3.194888178913738, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 4000 + }, + { + "epoch": 3.1956869009584663, + "grad_norm": 0.35546875, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 4001 + }, + { + "epoch": 3.196485623003195, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 4002 + }, + { + "epoch": 3.1972843450479234, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 4003 + }, + { + "epoch": 3.198083067092652, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 4004 + }, + { + "epoch": 3.1988817891373804, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 4005 + }, + { + "epoch": 3.1996805111821085, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 4006 + }, + { + "epoch": 3.200479233226837, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 4007 + }, + { + "epoch": 3.2012779552715656, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 4008 + }, + { + "epoch": 3.202076677316294, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 4009 + }, + { + "epoch": 3.202875399361022, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 4010 + }, + { + "epoch": 3.2036741214057507, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 4011 + }, + { + "epoch": 3.2044728434504792, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 4012 + }, + { + "epoch": 3.2052715654952078, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1317, + "step": 4013 + }, + { + "epoch": 3.2060702875399363, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 4014 + }, + { + "epoch": 3.2068690095846644, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 4015 + }, + { + "epoch": 3.207667731629393, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 4016 + }, + { + "epoch": 3.2084664536741214, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 4017 + }, + { + "epoch": 3.20926517571885, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 4018 + }, + { + "epoch": 3.2100638977635785, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 4019 + }, + { + "epoch": 3.2108626198083066, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 4020 + }, + { + "epoch": 3.211661341853035, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 4021 + }, + { + "epoch": 3.2124600638977636, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 4022 + }, + { + "epoch": 3.213258785942492, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 4023 + }, + { + "epoch": 3.2140575079872207, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 4024 + }, + { + "epoch": 3.2148562300319488, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 4025 + }, + { + "epoch": 3.2156549520766773, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 4026 + }, + { + "epoch": 3.216453674121406, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 4027 + }, + { + "epoch": 3.2172523961661343, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 4028 + }, + { + "epoch": 3.2180511182108624, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 4029 + }, + { + "epoch": 3.218849840255591, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 4030 + }, + { + "epoch": 3.2196485623003195, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 4031 + }, + { + "epoch": 3.220447284345048, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1364, + "step": 4032 + }, + { + "epoch": 3.2212460063897765, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 4033 + }, + { + "epoch": 3.2220447284345046, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 4034 + }, + { + "epoch": 3.222843450479233, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 4035 + }, + { + "epoch": 3.2236421725239617, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 4036 + }, + { + "epoch": 3.22444089456869, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4037 + }, + { + "epoch": 3.2252396166134187, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 4038 + }, + { + "epoch": 3.226038338658147, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 4039 + }, + { + "epoch": 3.2268370607028753, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 4040 + }, + { + "epoch": 3.227635782747604, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 4041 + }, + { + "epoch": 3.2284345047923324, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 4042 + }, + { + "epoch": 3.229233226837061, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 4043 + }, + { + "epoch": 3.230031948881789, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 4044 + }, + { + "epoch": 3.2308306709265175, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 4045 + }, + { + "epoch": 3.231629392971246, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 4046 + }, + { + "epoch": 3.2324281150159746, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 4047 + }, + { + "epoch": 3.2332268370607027, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 4048 + }, + { + "epoch": 3.234025559105431, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 4049 + }, + { + "epoch": 3.2348242811501597, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 4050 + }, + { + "epoch": 3.2356230031948883, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 4051 + }, + { + "epoch": 3.236421725239617, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 4052 + }, + { + "epoch": 3.237220447284345, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 4053 + }, + { + "epoch": 3.2380191693290734, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 4054 + }, + { + "epoch": 3.238817891373802, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 4055 + }, + { + "epoch": 3.2396166134185305, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4056 + }, + { + "epoch": 3.2404153354632586, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 4057 + }, + { + "epoch": 3.241214057507987, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 4058 + }, + { + "epoch": 3.2420127795527156, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4059 + }, + { + "epoch": 3.242811501597444, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 4060 + }, + { + "epoch": 3.2436102236421727, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 4061 + }, + { + "epoch": 3.244408945686901, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 4062 + }, + { + "epoch": 3.2452076677316293, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 4063 + }, + { + "epoch": 3.246006389776358, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 4064 + }, + { + "epoch": 3.2468051118210863, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 4065 + }, + { + "epoch": 3.247603833865815, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 4066 + }, + { + "epoch": 3.248402555910543, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 4067 + }, + { + "epoch": 3.2492012779552715, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 4068 + }, + { + "epoch": 3.25, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 4069 + }, + { + "epoch": 3.2507987220447285, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 4070 + }, + { + "epoch": 3.251597444089457, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 4071 + }, + { + "epoch": 3.252396166134185, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 4072 + }, + { + "epoch": 3.2531948881789137, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 4073 + }, + { + "epoch": 3.253993610223642, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 4074 + }, + { + "epoch": 3.2547923322683707, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 4075 + }, + { + "epoch": 3.255591054313099, + "grad_norm": 0.365234375, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 4076 + }, + { + "epoch": 3.2563897763578273, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 4077 + }, + { + "epoch": 3.257188498402556, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 4078 + }, + { + "epoch": 3.2579872204472844, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 4079 + }, + { + "epoch": 3.258785942492013, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 4080 + }, + { + "epoch": 3.2595846645367414, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 4081 + }, + { + "epoch": 3.2603833865814695, + "grad_norm": 0.384765625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 4082 + }, + { + "epoch": 3.261182108626198, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 4083 + }, + { + "epoch": 3.2619808306709266, + "grad_norm": 0.421875, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 4084 + }, + { + "epoch": 3.262779552715655, + "grad_norm": 0.482421875, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 4085 + }, + { + "epoch": 3.263578274760383, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 4086 + }, + { + "epoch": 3.2643769968051117, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 4087 + }, + { + "epoch": 3.2651757188498403, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 4088 + }, + { + "epoch": 3.265974440894569, + "grad_norm": 0.427734375, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 4089 + }, + { + "epoch": 3.2667731629392973, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 4090 + }, + { + "epoch": 3.2675718849840254, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 4091 + }, + { + "epoch": 3.268370607028754, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 4092 + }, + { + "epoch": 3.2691693290734825, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 4093 + }, + { + "epoch": 3.269968051118211, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 4094 + }, + { + "epoch": 3.270766773162939, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 4095 + }, + { + "epoch": 3.2715654952076676, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 4096 + }, + { + "epoch": 3.272364217252396, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1248, + "step": 4097 + }, + { + "epoch": 3.2731629392971247, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 4098 + }, + { + "epoch": 3.273961661341853, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 4099 + }, + { + "epoch": 3.2747603833865817, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 4100 + }, + { + "epoch": 3.27555910543131, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 4101 + }, + { + "epoch": 3.2763578274760383, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 4102 + }, + { + "epoch": 3.277156549520767, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 4103 + }, + { + "epoch": 3.2779552715654954, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 4104 + }, + { + "epoch": 3.2787539936102235, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 4105 + }, + { + "epoch": 3.279552715654952, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 4106 + }, + { + "epoch": 3.2803514376996805, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 4107 + }, + { + "epoch": 3.281150159744409, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 4108 + }, + { + "epoch": 3.2819488817891376, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 4109 + }, + { + "epoch": 3.2827476038338657, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 4110 + }, + { + "epoch": 3.283546325878594, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 4111 + }, + { + "epoch": 3.2843450479233227, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 4112 + }, + { + "epoch": 3.2851437699680512, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 4113 + }, + { + "epoch": 3.2859424920127793, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 4114 + }, + { + "epoch": 3.286741214057508, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 4115 + }, + { + "epoch": 3.2875399361022364, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4116 + }, + { + "epoch": 3.288338658146965, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 4117 + }, + { + "epoch": 3.2891373801916934, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 4118 + }, + { + "epoch": 3.289936102236422, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 4119 + }, + { + "epoch": 3.29073482428115, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 4120 + }, + { + "epoch": 3.2915335463258786, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 4121 + }, + { + "epoch": 3.292332268370607, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1293, + "step": 4122 + }, + { + "epoch": 3.2931309904153356, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 4123 + }, + { + "epoch": 3.2939297124600637, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 4124 + }, + { + "epoch": 3.2947284345047922, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 4125 + }, + { + "epoch": 3.2955271565495208, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 4126 + }, + { + "epoch": 3.2963258785942493, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 4127 + }, + { + "epoch": 3.297124600638978, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 4128 + }, + { + "epoch": 3.297923322683706, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 4129 + }, + { + "epoch": 3.2987220447284344, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 4130 + }, + { + "epoch": 3.299520766773163, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4131 + }, + { + "epoch": 3.3003194888178915, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 4132 + }, + { + "epoch": 3.3011182108626196, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 4133 + }, + { + "epoch": 3.301916932907348, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4134 + }, + { + "epoch": 3.3027156549520766, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 4135 + }, + { + "epoch": 3.303514376996805, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1273, + "step": 4136 + }, + { + "epoch": 3.3043130990415337, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1265, + "step": 4137 + }, + { + "epoch": 3.3051118210862622, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 4138 + }, + { + "epoch": 3.3059105431309903, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 4139 + }, + { + "epoch": 3.306709265175719, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 4140 + }, + { + "epoch": 3.3075079872204474, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.129, + "step": 4141 + }, + { + "epoch": 3.308306709265176, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 4142 + }, + { + "epoch": 3.309105431309904, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 4143 + }, + { + "epoch": 3.3099041533546325, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 4144 + }, + { + "epoch": 3.310702875399361, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4145 + }, + { + "epoch": 3.3115015974440896, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 4146 + }, + { + "epoch": 3.312300319488818, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 4147 + }, + { + "epoch": 3.313099041533546, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 4148 + }, + { + "epoch": 3.3138977635782747, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 4149 + }, + { + "epoch": 3.3146964856230032, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 4150 + }, + { + "epoch": 3.3154952076677318, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 4151 + }, + { + "epoch": 3.31629392971246, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 4152 + }, + { + "epoch": 3.3170926517571884, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1296, + "step": 4153 + }, + { + "epoch": 3.317891373801917, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 4154 + }, + { + "epoch": 3.3186900958466454, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 4155 + }, + { + "epoch": 3.319488817891374, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 4156 + }, + { + "epoch": 3.3202875399361025, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 4157 + }, + { + "epoch": 3.3210862619808306, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 4158 + }, + { + "epoch": 3.321884984025559, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 4159 + }, + { + "epoch": 3.3226837060702876, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 4160 + }, + { + "epoch": 3.323482428115016, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 4161 + }, + { + "epoch": 3.3242811501597442, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 4162 + }, + { + "epoch": 3.3250798722044728, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 4163 + }, + { + "epoch": 3.3258785942492013, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 4164 + }, + { + "epoch": 3.32667731629393, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 4165 + }, + { + "epoch": 3.3274760383386583, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 4166 + }, + { + "epoch": 3.3282747603833864, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 4167 + }, + { + "epoch": 3.329073482428115, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 4168 + }, + { + "epoch": 3.3298722044728435, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 4169 + }, + { + "epoch": 3.330670926517572, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 4170 + }, + { + "epoch": 3.3314696485623, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 4171 + }, + { + "epoch": 3.3322683706070286, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 4172 + }, + { + "epoch": 3.333067092651757, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1235, + "step": 4173 + }, + { + "epoch": 3.3338658146964857, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 4174 + }, + { + "epoch": 3.334664536741214, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 4175 + }, + { + "epoch": 3.3354632587859427, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 4176 + }, + { + "epoch": 3.336261980830671, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1341, + "step": 4177 + }, + { + "epoch": 3.3370607028753994, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 4178 + }, + { + "epoch": 3.337859424920128, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 4179 + }, + { + "epoch": 3.3386581469648564, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 4180 + }, + { + "epoch": 3.3394568690095845, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 4181 + }, + { + "epoch": 3.340255591054313, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 4182 + }, + { + "epoch": 3.3410543130990416, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 4183 + }, + { + "epoch": 3.34185303514377, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 4184 + }, + { + "epoch": 3.3426517571884986, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 4185 + }, + { + "epoch": 3.3434504792332267, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 4186 + }, + { + "epoch": 3.344249201277955, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 4187 + }, + { + "epoch": 3.3450479233226837, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 4188 + }, + { + "epoch": 3.3458466453674123, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 4189 + }, + { + "epoch": 3.3466453674121404, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 4190 + }, + { + "epoch": 3.347444089456869, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 4191 + }, + { + "epoch": 3.3482428115015974, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 4192 + }, + { + "epoch": 3.349041533546326, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 4193 + }, + { + "epoch": 3.3498402555910545, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 4194 + }, + { + "epoch": 3.3506389776357826, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 4195 + }, + { + "epoch": 3.351437699680511, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 4196 + }, + { + "epoch": 3.3522364217252396, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 4197 + }, + { + "epoch": 3.353035143769968, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 4198 + }, + { + "epoch": 3.3538338658146967, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 4199 + }, + { + "epoch": 3.3546325878594248, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 4200 + }, + { + "epoch": 3.3554313099041533, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 4201 + }, + { + "epoch": 3.356230031948882, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 4202 + }, + { + "epoch": 3.3570287539936103, + "grad_norm": 0.369140625, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4203 + }, + { + "epoch": 3.357827476038339, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 4204 + }, + { + "epoch": 3.358626198083067, + "grad_norm": 0.353515625, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 4205 + }, + { + "epoch": 3.3594249201277955, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 4206 + }, + { + "epoch": 3.360223642172524, + "grad_norm": 0.44140625, + "learning_rate": 0.0001, + "loss": 1.127, + "step": 4207 + }, + { + "epoch": 3.3610223642172525, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 4208 + }, + { + "epoch": 3.3618210862619806, + "grad_norm": 0.419921875, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 4209 + }, + { + "epoch": 3.362619808306709, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 4210 + }, + { + "epoch": 3.3634185303514377, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 4211 + }, + { + "epoch": 3.364217252396166, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 4212 + }, + { + "epoch": 3.3650159744408947, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4213 + }, + { + "epoch": 3.365814696485623, + "grad_norm": 0.33984375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 4214 + }, + { + "epoch": 3.3666134185303513, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 4215 + }, + { + "epoch": 3.36741214057508, + "grad_norm": 0.384765625, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 4216 + }, + { + "epoch": 3.3682108626198084, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 4217 + }, + { + "epoch": 3.369009584664537, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 4218 + }, + { + "epoch": 3.369808306709265, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 4219 + }, + { + "epoch": 3.3706070287539935, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 4220 + }, + { + "epoch": 3.371405750798722, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4221 + }, + { + "epoch": 3.3722044728434506, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 4222 + }, + { + "epoch": 3.373003194888179, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 4223 + }, + { + "epoch": 3.373801916932907, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 4224 + }, + { + "epoch": 3.3746006389776357, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 4225 + }, + { + "epoch": 3.3753993610223643, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 4226 + }, + { + "epoch": 3.376198083067093, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 4227 + }, + { + "epoch": 3.376996805111821, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 4228 + }, + { + "epoch": 3.3777955271565494, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 4229 + }, + { + "epoch": 3.378594249201278, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 4230 + }, + { + "epoch": 3.3793929712460065, + "grad_norm": 0.458984375, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 4231 + }, + { + "epoch": 3.380191693290735, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 4232 + }, + { + "epoch": 3.380990415335463, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.1287, + "step": 4233 + }, + { + "epoch": 3.3817891373801916, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 4234 + }, + { + "epoch": 3.38258785942492, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 4235 + }, + { + "epoch": 3.3833865814696487, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 4236 + }, + { + "epoch": 3.384185303514377, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 4237 + }, + { + "epoch": 3.3849840255591053, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 4238 + }, + { + "epoch": 3.385782747603834, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 4239 + }, + { + "epoch": 3.3865814696485623, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 4240 + }, + { + "epoch": 3.387380191693291, + "grad_norm": 0.44921875, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 4241 + }, + { + "epoch": 3.3881789137380194, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 4242 + }, + { + "epoch": 3.3889776357827475, + "grad_norm": 0.40234375, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 4243 + }, + { + "epoch": 3.389776357827476, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 4244 + }, + { + "epoch": 3.3905750798722045, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 4245 + }, + { + "epoch": 3.391373801916933, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 4246 + }, + { + "epoch": 3.392172523961661, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1273, + "step": 4247 + }, + { + "epoch": 3.3929712460063897, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 4248 + }, + { + "epoch": 3.393769968051118, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1305, + "step": 4249 + }, + { + "epoch": 3.3945686900958467, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 4250 + }, + { + "epoch": 3.3953674121405752, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 4251 + }, + { + "epoch": 3.3961661341853033, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 4252 + }, + { + "epoch": 3.396964856230032, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 4253 + }, + { + "epoch": 3.3977635782747604, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 4254 + }, + { + "epoch": 3.398562300319489, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 4255 + }, + { + "epoch": 3.3993610223642174, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 4256 + }, + { + "epoch": 3.4001597444089455, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 4257 + }, + { + "epoch": 3.400958466453674, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 4258 + }, + { + "epoch": 3.4017571884984026, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 4259 + }, + { + "epoch": 3.402555910543131, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 4260 + }, + { + "epoch": 3.4033546325878596, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 4261 + }, + { + "epoch": 3.4041533546325877, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 4262 + }, + { + "epoch": 3.4049520766773163, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 4263 + }, + { + "epoch": 3.405750798722045, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4264 + }, + { + "epoch": 3.4065495207667733, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 4265 + }, + { + "epoch": 3.4073482428115014, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 4266 + }, + { + "epoch": 3.40814696485623, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 4267 + }, + { + "epoch": 3.4089456869009584, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 4268 + }, + { + "epoch": 3.409744408945687, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 4269 + }, + { + "epoch": 3.4105431309904155, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 4270 + }, + { + "epoch": 3.4113418530351436, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1264, + "step": 4271 + }, + { + "epoch": 3.412140575079872, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 4272 + }, + { + "epoch": 3.4129392971246006, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 4273 + }, + { + "epoch": 3.413738019169329, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 4274 + }, + { + "epoch": 3.4145367412140573, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 4275 + }, + { + "epoch": 3.415335463258786, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 4276 + }, + { + "epoch": 3.4161341853035143, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 4277 + }, + { + "epoch": 3.416932907348243, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 4278 + }, + { + "epoch": 3.4177316293929714, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 4279 + }, + { + "epoch": 3.4185303514377, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1295, + "step": 4280 + }, + { + "epoch": 3.419329073482428, + "grad_norm": 0.3828125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 4281 + }, + { + "epoch": 3.4201277955271565, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 4282 + }, + { + "epoch": 3.420926517571885, + "grad_norm": 0.3984375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 4283 + }, + { + "epoch": 3.4217252396166136, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 4284 + }, + { + "epoch": 3.4225239616613417, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 4285 + }, + { + "epoch": 3.42332268370607, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 4286 + }, + { + "epoch": 3.4241214057507987, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 4287 + }, + { + "epoch": 3.4249201277955272, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 4288 + }, + { + "epoch": 3.4257188498402558, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 4289 + }, + { + "epoch": 3.426517571884984, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 4290 + }, + { + "epoch": 3.4273162939297124, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 4291 + }, + { + "epoch": 3.428115015974441, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 4292 + }, + { + "epoch": 3.4289137380191694, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 4293 + }, + { + "epoch": 3.4297124600638975, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 4294 + }, + { + "epoch": 3.430511182108626, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 4295 + }, + { + "epoch": 3.4313099041533546, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 4296 + }, + { + "epoch": 3.432108626198083, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4297 + }, + { + "epoch": 3.4329073482428116, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 4298 + }, + { + "epoch": 3.43370607028754, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 4299 + }, + { + "epoch": 3.4345047923322682, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 4300 + }, + { + "epoch": 3.4353035143769968, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 4301 + }, + { + "epoch": 3.4361022364217253, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 4302 + }, + { + "epoch": 3.436900958466454, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 4303 + }, + { + "epoch": 3.437699680511182, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1286, + "step": 4304 + }, + { + "epoch": 3.4384984025559104, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 4305 + }, + { + "epoch": 3.439297124600639, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 4306 + }, + { + "epoch": 3.4400958466453675, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 4307 + }, + { + "epoch": 3.440894568690096, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 4308 + }, + { + "epoch": 3.441693290734824, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 4309 + }, + { + "epoch": 3.4424920127795526, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 4310 + }, + { + "epoch": 3.443290734824281, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 4311 + }, + { + "epoch": 3.4440894568690097, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 4312 + }, + { + "epoch": 3.4448881789137378, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 4313 + }, + { + "epoch": 3.4456869009584663, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 4314 + }, + { + "epoch": 3.446485623003195, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 4315 + }, + { + "epoch": 3.4472843450479234, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 4316 + }, + { + "epoch": 3.448083067092652, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 4317 + }, + { + "epoch": 3.4488817891373804, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 4318 + }, + { + "epoch": 3.4496805111821085, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1282, + "step": 4319 + }, + { + "epoch": 3.450479233226837, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 4320 + }, + { + "epoch": 3.4512779552715656, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 4321 + }, + { + "epoch": 3.452076677316294, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 4322 + }, + { + "epoch": 3.452875399361022, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 4323 + }, + { + "epoch": 3.4536741214057507, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 4324 + }, + { + "epoch": 3.4544728434504792, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 4325 + }, + { + "epoch": 3.4552715654952078, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 4326 + }, + { + "epoch": 3.4560702875399363, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 4327 + }, + { + "epoch": 3.4568690095846644, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 4328 + }, + { + "epoch": 3.457667731629393, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1294, + "step": 4329 + }, + { + "epoch": 3.4584664536741214, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 4330 + }, + { + "epoch": 3.45926517571885, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 4331 + }, + { + "epoch": 3.460063897763578, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 4332 + }, + { + "epoch": 3.4608626198083066, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4333 + }, + { + "epoch": 3.461661341853035, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 4334 + }, + { + "epoch": 3.4624600638977636, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 4335 + }, + { + "epoch": 3.463258785942492, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 4336 + }, + { + "epoch": 3.4640575079872207, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1281, + "step": 4337 + }, + { + "epoch": 3.4648562300319488, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 4338 + }, + { + "epoch": 3.4656549520766773, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 4339 + }, + { + "epoch": 3.466453674121406, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 4340 + }, + { + "epoch": 3.4672523961661343, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 4341 + }, + { + "epoch": 3.4680511182108624, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 4342 + }, + { + "epoch": 3.468849840255591, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4343 + }, + { + "epoch": 3.4696485623003195, + "grad_norm": 0.412109375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 4344 + }, + { + "epoch": 3.470447284345048, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 4345 + }, + { + "epoch": 3.4712460063897765, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 4346 + }, + { + "epoch": 3.4720447284345046, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 4347 + }, + { + "epoch": 3.472843450479233, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 4348 + }, + { + "epoch": 3.4736421725239617, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 4349 + }, + { + "epoch": 3.47444089456869, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 4350 + }, + { + "epoch": 3.4752396166134183, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 4351 + }, + { + "epoch": 3.476038338658147, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 4352 + }, + { + "epoch": 3.4768370607028753, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 4353 + }, + { + "epoch": 3.477635782747604, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 4354 + }, + { + "epoch": 3.4784345047923324, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 4355 + }, + { + "epoch": 3.479233226837061, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 4356 + }, + { + "epoch": 3.480031948881789, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 4357 + }, + { + "epoch": 3.4808306709265175, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 4358 + }, + { + "epoch": 3.481629392971246, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 4359 + }, + { + "epoch": 3.4824281150159746, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 4360 + }, + { + "epoch": 3.4832268370607027, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 4361 + }, + { + "epoch": 3.484025559105431, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 4362 + }, + { + "epoch": 3.4848242811501597, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 4363 + }, + { + "epoch": 3.4856230031948883, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 4364 + }, + { + "epoch": 3.486421725239617, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1309, + "step": 4365 + }, + { + "epoch": 3.487220447284345, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 4366 + }, + { + "epoch": 3.4880191693290734, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 4367 + }, + { + "epoch": 3.488817891373802, + "grad_norm": 0.412109375, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 4368 + }, + { + "epoch": 3.4896166134185305, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4369 + }, + { + "epoch": 3.4904153354632586, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 4370 + }, + { + "epoch": 3.491214057507987, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 4371 + }, + { + "epoch": 3.4920127795527156, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 4372 + }, + { + "epoch": 3.492811501597444, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 4373 + }, + { + "epoch": 3.4936102236421727, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 4374 + }, + { + "epoch": 3.494408945686901, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 4375 + }, + { + "epoch": 3.4952076677316293, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 4376 + }, + { + "epoch": 3.496006389776358, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 4377 + }, + { + "epoch": 3.4968051118210863, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 4378 + }, + { + "epoch": 3.497603833865815, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 4379 + }, + { + "epoch": 3.498402555910543, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 4380 + }, + { + "epoch": 3.4992012779552715, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 4381 + }, + { + "epoch": 3.5, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 4382 + }, + { + "epoch": 3.5007987220447285, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 4383 + }, + { + "epoch": 3.501597444089457, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1325, + "step": 4384 + }, + { + "epoch": 3.502396166134185, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1324, + "step": 4385 + }, + { + "epoch": 3.5031948881789137, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 4386 + }, + { + "epoch": 3.503993610223642, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 4387 + }, + { + "epoch": 3.5047923322683707, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 4388 + }, + { + "epoch": 3.505591054313099, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 4389 + }, + { + "epoch": 3.5063897763578273, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 4390 + }, + { + "epoch": 3.507188498402556, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 4391 + }, + { + "epoch": 3.5079872204472844, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 4392 + }, + { + "epoch": 3.508785942492013, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 4393 + }, + { + "epoch": 3.5095846645367414, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 4394 + }, + { + "epoch": 3.5103833865814695, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 4395 + }, + { + "epoch": 3.511182108626198, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 4396 + }, + { + "epoch": 3.5119808306709266, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 4397 + }, + { + "epoch": 3.512779552715655, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 4398 + }, + { + "epoch": 3.513578274760383, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 4399 + }, + { + "epoch": 3.5143769968051117, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 4400 + }, + { + "epoch": 3.5151757188498403, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 4401 + }, + { + "epoch": 3.515974440894569, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 4402 + }, + { + "epoch": 3.5167731629392973, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 4403 + }, + { + "epoch": 3.5175718849840254, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 4404 + }, + { + "epoch": 3.518370607028754, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 4405 + }, + { + "epoch": 3.5191693290734825, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 4406 + }, + { + "epoch": 3.519968051118211, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 4407 + }, + { + "epoch": 3.520766773162939, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 4408 + }, + { + "epoch": 3.5215654952076676, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 4409 + }, + { + "epoch": 3.522364217252396, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 4410 + }, + { + "epoch": 3.5231629392971247, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 4411 + }, + { + "epoch": 3.523961661341853, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 4412 + }, + { + "epoch": 3.5247603833865817, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 4413 + }, + { + "epoch": 3.52555910543131, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1314, + "step": 4414 + }, + { + "epoch": 3.5263578274760383, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 4415 + }, + { + "epoch": 3.527156549520767, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 4416 + }, + { + "epoch": 3.527955271565495, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 4417 + }, + { + "epoch": 3.5287539936102235, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 4418 + }, + { + "epoch": 3.529552715654952, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 4419 + }, + { + "epoch": 3.5303514376996805, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4420 + }, + { + "epoch": 3.531150159744409, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 4421 + }, + { + "epoch": 3.5319488817891376, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 4422 + }, + { + "epoch": 3.5327476038338657, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 4423 + }, + { + "epoch": 3.533546325878594, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 4424 + }, + { + "epoch": 3.5343450479233227, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 4425 + }, + { + "epoch": 3.5351437699680512, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 4426 + }, + { + "epoch": 3.5359424920127793, + "grad_norm": 0.09716796875, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 4427 + }, + { + "epoch": 3.536741214057508, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 4428 + }, + { + "epoch": 3.5375399361022364, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 4429 + }, + { + "epoch": 3.538338658146965, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 4430 + }, + { + "epoch": 3.5391373801916934, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 4431 + }, + { + "epoch": 3.539936102236422, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 4432 + }, + { + "epoch": 3.54073482428115, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 4433 + }, + { + "epoch": 3.5415335463258786, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 4434 + }, + { + "epoch": 3.542332268370607, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1258, + "step": 4435 + }, + { + "epoch": 3.543130990415335, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 4436 + }, + { + "epoch": 3.5439297124600637, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 4437 + }, + { + "epoch": 3.5447284345047922, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 4438 + }, + { + "epoch": 3.5455271565495208, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 4439 + }, + { + "epoch": 3.5463258785942493, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 4440 + }, + { + "epoch": 3.547124600638978, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 4441 + }, + { + "epoch": 3.547923322683706, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 4442 + }, + { + "epoch": 3.5487220447284344, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 4443 + }, + { + "epoch": 3.549520766773163, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 4444 + }, + { + "epoch": 3.5503194888178915, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 4445 + }, + { + "epoch": 3.5511182108626196, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 4446 + }, + { + "epoch": 3.551916932907348, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 4447 + }, + { + "epoch": 3.5527156549520766, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1248, + "step": 4448 + }, + { + "epoch": 3.553514376996805, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 4449 + }, + { + "epoch": 3.5543130990415337, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 4450 + }, + { + "epoch": 3.5551118210862622, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 4451 + }, + { + "epoch": 3.5559105431309903, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1325, + "step": 4452 + }, + { + "epoch": 3.556709265175719, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1326, + "step": 4453 + }, + { + "epoch": 3.5575079872204474, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 4454 + }, + { + "epoch": 3.5583067092651754, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 4455 + }, + { + "epoch": 3.559105431309904, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 4456 + }, + { + "epoch": 3.5599041533546325, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 4457 + }, + { + "epoch": 3.560702875399361, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 4458 + }, + { + "epoch": 3.5615015974440896, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 4459 + }, + { + "epoch": 3.562300319488818, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 4460 + }, + { + "epoch": 3.563099041533546, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 4461 + }, + { + "epoch": 3.5638977635782747, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 4462 + }, + { + "epoch": 3.5646964856230032, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 4463 + }, + { + "epoch": 3.5654952076677318, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 4464 + }, + { + "epoch": 3.56629392971246, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 4465 + }, + { + "epoch": 3.5670926517571884, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4466 + }, + { + "epoch": 3.567891373801917, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 4467 + }, + { + "epoch": 3.5686900958466454, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 4468 + }, + { + "epoch": 3.569488817891374, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 4469 + }, + { + "epoch": 3.5702875399361025, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 4470 + }, + { + "epoch": 3.5710862619808306, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 4471 + }, + { + "epoch": 3.571884984025559, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 4472 + }, + { + "epoch": 3.5726837060702876, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 4473 + }, + { + "epoch": 3.5734824281150157, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4474 + }, + { + "epoch": 3.5742811501597442, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 4475 + }, + { + "epoch": 3.5750798722044728, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4476 + }, + { + "epoch": 3.5758785942492013, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 4477 + }, + { + "epoch": 3.57667731629393, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 4478 + }, + { + "epoch": 3.5774760383386583, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 4479 + }, + { + "epoch": 3.5782747603833864, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 4480 + }, + { + "epoch": 3.579073482428115, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 4481 + }, + { + "epoch": 3.5798722044728435, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 4482 + }, + { + "epoch": 3.580670926517572, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 4483 + }, + { + "epoch": 3.5814696485623, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 4484 + }, + { + "epoch": 3.5822683706070286, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 4485 + }, + { + "epoch": 3.583067092651757, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 4486 + }, + { + "epoch": 3.5838658146964857, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 4487 + }, + { + "epoch": 3.584664536741214, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1235, + "step": 4488 + }, + { + "epoch": 3.5854632587859427, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 4489 + }, + { + "epoch": 3.586261980830671, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 4490 + }, + { + "epoch": 3.5870607028753994, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 4491 + }, + { + "epoch": 3.587859424920128, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 4492 + }, + { + "epoch": 3.588658146964856, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 4493 + }, + { + "epoch": 3.5894568690095845, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 4494 + }, + { + "epoch": 3.590255591054313, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 4495 + }, + { + "epoch": 3.5910543130990416, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 4496 + }, + { + "epoch": 3.59185303514377, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 4497 + }, + { + "epoch": 3.5926517571884986, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 4498 + }, + { + "epoch": 3.5934504792332267, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 4499 + }, + { + "epoch": 3.594249201277955, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1256, + "step": 4500 + }, + { + "epoch": 3.5950479233226837, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 4501 + }, + { + "epoch": 3.5958466453674123, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 4502 + }, + { + "epoch": 3.5966453674121404, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 4503 + }, + { + "epoch": 3.597444089456869, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 4504 + }, + { + "epoch": 3.5982428115015974, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 4505 + }, + { + "epoch": 3.599041533546326, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 4506 + }, + { + "epoch": 3.5998402555910545, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 4507 + }, + { + "epoch": 3.600638977635783, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 4508 + }, + { + "epoch": 3.601437699680511, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 4509 + }, + { + "epoch": 3.6022364217252396, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 4510 + }, + { + "epoch": 3.603035143769968, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 4511 + }, + { + "epoch": 3.6038338658146962, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 4512 + }, + { + "epoch": 3.6046325878594248, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 4513 + }, + { + "epoch": 3.6054313099041533, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 4514 + }, + { + "epoch": 3.606230031948882, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 4515 + }, + { + "epoch": 3.6070287539936103, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 4516 + }, + { + "epoch": 3.607827476038339, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 4517 + }, + { + "epoch": 3.608626198083067, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 4518 + }, + { + "epoch": 3.6094249201277955, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 4519 + }, + { + "epoch": 3.610223642172524, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 4520 + }, + { + "epoch": 3.6110223642172525, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 4521 + }, + { + "epoch": 3.6118210862619806, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 4522 + }, + { + "epoch": 3.612619808306709, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 4523 + }, + { + "epoch": 3.6134185303514377, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 4524 + }, + { + "epoch": 3.614217252396166, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 4525 + }, + { + "epoch": 3.6150159744408947, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 4526 + }, + { + "epoch": 3.6158146964856233, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4527 + }, + { + "epoch": 3.6166134185303513, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 4528 + }, + { + "epoch": 3.61741214057508, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 4529 + }, + { + "epoch": 3.6182108626198084, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 4530 + }, + { + "epoch": 3.6190095846645365, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 4531 + }, + { + "epoch": 3.619808306709265, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 4532 + }, + { + "epoch": 3.6206070287539935, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 4533 + }, + { + "epoch": 3.621405750798722, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 4534 + }, + { + "epoch": 3.6222044728434506, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 4535 + }, + { + "epoch": 3.623003194888179, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 4536 + }, + { + "epoch": 3.623801916932907, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 4537 + }, + { + "epoch": 3.6246006389776357, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4538 + }, + { + "epoch": 3.6253993610223643, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 4539 + }, + { + "epoch": 3.626198083067093, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 4540 + }, + { + "epoch": 3.626996805111821, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 4541 + }, + { + "epoch": 3.6277955271565494, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 4542 + }, + { + "epoch": 3.628594249201278, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 4543 + }, + { + "epoch": 3.6293929712460065, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 4544 + }, + { + "epoch": 3.630191693290735, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 4545 + }, + { + "epoch": 3.6309904153354635, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 4546 + }, + { + "epoch": 3.6317891373801916, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 4547 + }, + { + "epoch": 3.63258785942492, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 4548 + }, + { + "epoch": 3.6333865814696487, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 4549 + }, + { + "epoch": 3.6341853035143767, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 4550 + }, + { + "epoch": 3.6349840255591053, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 4551 + }, + { + "epoch": 3.635782747603834, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 4552 + }, + { + "epoch": 3.6365814696485623, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 4553 + }, + { + "epoch": 3.637380191693291, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 4554 + }, + { + "epoch": 3.6381789137380194, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 4555 + }, + { + "epoch": 3.6389776357827475, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 4556 + }, + { + "epoch": 3.639776357827476, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 4557 + }, + { + "epoch": 3.6405750798722045, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 4558 + }, + { + "epoch": 3.641373801916933, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 4559 + }, + { + "epoch": 3.642172523961661, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 4560 + }, + { + "epoch": 3.6429712460063897, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 4561 + }, + { + "epoch": 3.643769968051118, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 4562 + }, + { + "epoch": 3.6445686900958467, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 4563 + }, + { + "epoch": 3.6453674121405752, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 4564 + }, + { + "epoch": 3.6461661341853038, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 4565 + }, + { + "epoch": 3.646964856230032, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 4566 + }, + { + "epoch": 3.6477635782747604, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 4567 + }, + { + "epoch": 3.648562300319489, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 4568 + }, + { + "epoch": 3.649361022364217, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 4569 + }, + { + "epoch": 3.6501597444089455, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 4570 + }, + { + "epoch": 3.650958466453674, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 4571 + }, + { + "epoch": 3.6517571884984026, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 4572 + }, + { + "epoch": 3.652555910543131, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 4573 + }, + { + "epoch": 3.6533546325878596, + "grad_norm": 0.396484375, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 4574 + }, + { + "epoch": 3.6541533546325877, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 4575 + }, + { + "epoch": 3.6549520766773163, + "grad_norm": 0.5546875, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 4576 + }, + { + "epoch": 3.655750798722045, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 4577 + }, + { + "epoch": 3.6565495207667733, + "grad_norm": 0.359375, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 4578 + }, + { + "epoch": 3.6573482428115014, + "grad_norm": 0.478515625, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 4579 + }, + { + "epoch": 3.65814696485623, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 4580 + }, + { + "epoch": 3.6589456869009584, + "grad_norm": 0.8359375, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 4581 + }, + { + "epoch": 3.659744408945687, + "grad_norm": 0.77734375, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 4582 + }, + { + "epoch": 3.6605431309904155, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1304, + "step": 4583 + }, + { + "epoch": 3.661341853035144, + "grad_norm": 0.462890625, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 4584 + }, + { + "epoch": 3.662140575079872, + "grad_norm": 0.376953125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 4585 + }, + { + "epoch": 3.6629392971246006, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 4586 + }, + { + "epoch": 3.663738019169329, + "grad_norm": 0.6796875, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 4587 + }, + { + "epoch": 3.6645367412140573, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 4588 + }, + { + "epoch": 3.665335463258786, + "grad_norm": 0.5078125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 4589 + }, + { + "epoch": 3.6661341853035143, + "grad_norm": 0.388671875, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 4590 + }, + { + "epoch": 3.666932907348243, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4591 + }, + { + "epoch": 3.6677316293929714, + "grad_norm": 0.73046875, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 4592 + }, + { + "epoch": 3.6685303514377, + "grad_norm": 0.4765625, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 4593 + }, + { + "epoch": 3.669329073482428, + "grad_norm": 0.3828125, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 4594 + }, + { + "epoch": 3.6701277955271565, + "grad_norm": 1.15625, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 4595 + }, + { + "epoch": 3.670926517571885, + "grad_norm": 0.9296875, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 4596 + }, + { + "epoch": 3.6717252396166136, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 4597 + }, + { + "epoch": 3.6725239616613417, + "grad_norm": 0.76171875, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 4598 + }, + { + "epoch": 3.67332268370607, + "grad_norm": 0.87109375, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 4599 + }, + { + "epoch": 3.6741214057507987, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 4600 + }, + { + "epoch": 3.6749201277955272, + "grad_norm": 0.484375, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 4601 + }, + { + "epoch": 3.6757188498402558, + "grad_norm": 0.42578125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 4602 + }, + { + "epoch": 3.6765175718849843, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 4603 + }, + { + "epoch": 3.6773162939297124, + "grad_norm": 0.435546875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 4604 + }, + { + "epoch": 3.678115015974441, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 4605 + }, + { + "epoch": 3.6789137380191694, + "grad_norm": 0.609375, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 4606 + }, + { + "epoch": 3.6797124600638975, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 4607 + }, + { + "epoch": 3.680511182108626, + "grad_norm": 0.46484375, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 4608 + }, + { + "epoch": 3.6813099041533546, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 4609 + }, + { + "epoch": 3.682108626198083, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 4610 + }, + { + "epoch": 3.6829073482428116, + "grad_norm": 0.59765625, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 4611 + }, + { + "epoch": 3.68370607028754, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 4612 + }, + { + "epoch": 3.6845047923322682, + "grad_norm": 0.359375, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 4613 + }, + { + "epoch": 3.6853035143769968, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 4614 + }, + { + "epoch": 3.6861022364217253, + "grad_norm": 0.357421875, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 4615 + }, + { + "epoch": 3.686900958466454, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 4616 + }, + { + "epoch": 3.687699680511182, + "grad_norm": 0.470703125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 4617 + }, + { + "epoch": 3.6884984025559104, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 4618 + }, + { + "epoch": 3.689297124600639, + "grad_norm": 0.400390625, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 4619 + }, + { + "epoch": 3.6900958466453675, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 4620 + }, + { + "epoch": 3.690894568690096, + "grad_norm": 0.443359375, + "learning_rate": 0.0001, + "loss": 1.1266, + "step": 4621 + }, + { + "epoch": 3.6916932907348246, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 4622 + }, + { + "epoch": 3.6924920127795526, + "grad_norm": 0.5390625, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4623 + }, + { + "epoch": 3.693290734824281, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 4624 + }, + { + "epoch": 3.6940894568690097, + "grad_norm": 0.51171875, + "learning_rate": 0.0001, + "loss": 1.1235, + "step": 4625 + }, + { + "epoch": 3.6948881789137378, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 4626 + }, + { + "epoch": 3.6956869009584663, + "grad_norm": 0.490234375, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 4627 + }, + { + "epoch": 3.696485623003195, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 4628 + }, + { + "epoch": 3.6972843450479234, + "grad_norm": 0.390625, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 4629 + }, + { + "epoch": 3.698083067092652, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 4630 + }, + { + "epoch": 3.6988817891373804, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4631 + }, + { + "epoch": 3.6996805111821085, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4632 + }, + { + "epoch": 3.700479233226837, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 4633 + }, + { + "epoch": 3.7012779552715656, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1321, + "step": 4634 + }, + { + "epoch": 3.702076677316294, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4635 + }, + { + "epoch": 3.702875399361022, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 4636 + }, + { + "epoch": 3.7036741214057507, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 4637 + }, + { + "epoch": 3.7044728434504792, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 4638 + }, + { + "epoch": 3.7052715654952078, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4639 + }, + { + "epoch": 3.7060702875399363, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 4640 + }, + { + "epoch": 3.706869009584665, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 4641 + }, + { + "epoch": 3.707667731629393, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 4642 + }, + { + "epoch": 3.7084664536741214, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 4643 + }, + { + "epoch": 3.70926517571885, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 4644 + }, + { + "epoch": 3.710063897763578, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 4645 + }, + { + "epoch": 3.7108626198083066, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 4646 + }, + { + "epoch": 3.711661341853035, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 4647 + }, + { + "epoch": 3.7124600638977636, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1274, + "step": 4648 + }, + { + "epoch": 3.713258785942492, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 4649 + }, + { + "epoch": 3.7140575079872207, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 4650 + }, + { + "epoch": 3.7148562300319488, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 4651 + }, + { + "epoch": 3.7156549520766773, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 4652 + }, + { + "epoch": 3.716453674121406, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 4653 + }, + { + "epoch": 3.7172523961661343, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 4654 + }, + { + "epoch": 3.7180511182108624, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 4655 + }, + { + "epoch": 3.718849840255591, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 4656 + }, + { + "epoch": 3.7196485623003195, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 4657 + }, + { + "epoch": 3.720447284345048, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1235, + "step": 4658 + }, + { + "epoch": 3.7212460063897765, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 4659 + }, + { + "epoch": 3.722044728434505, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 4660 + }, + { + "epoch": 3.722843450479233, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 4661 + }, + { + "epoch": 3.7236421725239617, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 4662 + }, + { + "epoch": 3.72444089456869, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 4663 + }, + { + "epoch": 3.7252396166134183, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 4664 + }, + { + "epoch": 3.726038338658147, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 4665 + }, + { + "epoch": 3.7268370607028753, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 4666 + }, + { + "epoch": 3.727635782747604, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 4667 + }, + { + "epoch": 3.7284345047923324, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 4668 + }, + { + "epoch": 3.729233226837061, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 4669 + }, + { + "epoch": 3.730031948881789, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 4670 + }, + { + "epoch": 3.7308306709265175, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 4671 + }, + { + "epoch": 3.731629392971246, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 4672 + }, + { + "epoch": 3.7324281150159746, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 4673 + }, + { + "epoch": 3.7332268370607027, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 4674 + }, + { + "epoch": 3.734025559105431, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 4675 + }, + { + "epoch": 3.7348242811501597, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 4676 + }, + { + "epoch": 3.7356230031948883, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 4677 + }, + { + "epoch": 3.736421725239617, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 4678 + }, + { + "epoch": 3.737220447284345, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 4679 + }, + { + "epoch": 3.7380191693290734, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 4680 + }, + { + "epoch": 3.738817891373802, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 4681 + }, + { + "epoch": 3.7396166134185305, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 4682 + }, + { + "epoch": 3.7404153354632586, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 4683 + }, + { + "epoch": 3.741214057507987, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 4684 + }, + { + "epoch": 3.7420127795527156, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 4685 + }, + { + "epoch": 3.742811501597444, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 4686 + }, + { + "epoch": 3.7436102236421727, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 4687 + }, + { + "epoch": 3.744408945686901, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 4688 + }, + { + "epoch": 3.7452076677316293, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 4689 + }, + { + "epoch": 3.746006389776358, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 4690 + }, + { + "epoch": 3.7468051118210863, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 4691 + }, + { + "epoch": 3.747603833865815, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 4692 + }, + { + "epoch": 3.748402555910543, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 4693 + }, + { + "epoch": 3.7492012779552715, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 4694 + }, + { + "epoch": 3.75, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 4695 + }, + { + "epoch": 3.7507987220447285, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 4696 + }, + { + "epoch": 3.751597444089457, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 4697 + }, + { + "epoch": 3.752396166134185, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 4698 + }, + { + "epoch": 3.7531948881789137, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 4699 + }, + { + "epoch": 3.753993610223642, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 4700 + }, + { + "epoch": 3.7547923322683707, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 4701 + }, + { + "epoch": 3.755591054313099, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 4702 + }, + { + "epoch": 3.7563897763578273, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 4703 + }, + { + "epoch": 3.757188498402556, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 4704 + }, + { + "epoch": 3.7579872204472844, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1283, + "step": 4705 + }, + { + "epoch": 3.758785942492013, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 4706 + }, + { + "epoch": 3.7595846645367414, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 4707 + }, + { + "epoch": 3.7603833865814695, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 4708 + }, + { + "epoch": 3.761182108626198, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 4709 + }, + { + "epoch": 3.7619808306709266, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1267, + "step": 4710 + }, + { + "epoch": 3.762779552715655, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4711 + }, + { + "epoch": 3.763578274760383, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 4712 + }, + { + "epoch": 3.7643769968051117, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 4713 + }, + { + "epoch": 3.7651757188498403, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 4714 + }, + { + "epoch": 3.765974440894569, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 4715 + }, + { + "epoch": 3.7667731629392973, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 4716 + }, + { + "epoch": 3.7675718849840254, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 4717 + }, + { + "epoch": 3.768370607028754, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 4718 + }, + { + "epoch": 3.7691693290734825, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 4719 + }, + { + "epoch": 3.769968051118211, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 4720 + }, + { + "epoch": 3.770766773162939, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4721 + }, + { + "epoch": 3.7715654952076676, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 4722 + }, + { + "epoch": 3.772364217252396, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 4723 + }, + { + "epoch": 3.7731629392971247, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1318, + "step": 4724 + }, + { + "epoch": 3.773961661341853, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 4725 + }, + { + "epoch": 3.7747603833865817, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4726 + }, + { + "epoch": 3.77555910543131, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 4727 + }, + { + "epoch": 3.7763578274760383, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 4728 + }, + { + "epoch": 3.777156549520767, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 4729 + }, + { + "epoch": 3.777955271565495, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 4730 + }, + { + "epoch": 3.7787539936102235, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 4731 + }, + { + "epoch": 3.779552715654952, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 4732 + }, + { + "epoch": 3.7803514376996805, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 4733 + }, + { + "epoch": 3.781150159744409, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 4734 + }, + { + "epoch": 3.7819488817891376, + "grad_norm": 0.34375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 4735 + }, + { + "epoch": 3.7827476038338657, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 4736 + }, + { + "epoch": 3.783546325878594, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 4737 + }, + { + "epoch": 3.7843450479233227, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 4738 + }, + { + "epoch": 3.7851437699680512, + "grad_norm": 0.40625, + "learning_rate": 0.0001, + "loss": 1.1288, + "step": 4739 + }, + { + "epoch": 3.7859424920127793, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 4740 + }, + { + "epoch": 3.786741214057508, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 4741 + }, + { + "epoch": 3.7875399361022364, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 4742 + }, + { + "epoch": 3.788338658146965, + "grad_norm": 0.37890625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 4743 + }, + { + "epoch": 3.7891373801916934, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 4744 + }, + { + "epoch": 3.789936102236422, + "grad_norm": 0.443359375, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 4745 + }, + { + "epoch": 3.79073482428115, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 4746 + }, + { + "epoch": 3.7915335463258786, + "grad_norm": 0.4375, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4747 + }, + { + "epoch": 3.792332268370607, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 4748 + }, + { + "epoch": 3.793130990415335, + "grad_norm": 0.462890625, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 4749 + }, + { + "epoch": 3.7939297124600637, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 4750 + }, + { + "epoch": 3.7947284345047922, + "grad_norm": 0.4921875, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 4751 + }, + { + "epoch": 3.7955271565495208, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 4752 + }, + { + "epoch": 3.7963258785942493, + "grad_norm": 0.419921875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 4753 + }, + { + "epoch": 3.797124600638978, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 4754 + }, + { + "epoch": 3.797923322683706, + "grad_norm": 0.455078125, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 4755 + }, + { + "epoch": 3.7987220447284344, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4756 + }, + { + "epoch": 3.799520766773163, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 4757 + }, + { + "epoch": 3.8003194888178915, + "grad_norm": 0.421875, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 4758 + }, + { + "epoch": 3.8011182108626196, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1311, + "step": 4759 + }, + { + "epoch": 3.801916932907348, + "grad_norm": 0.68359375, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 4760 + }, + { + "epoch": 3.8027156549520766, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 4761 + }, + { + "epoch": 3.803514376996805, + "grad_norm": 0.494140625, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 4762 + }, + { + "epoch": 3.8043130990415337, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 4763 + }, + { + "epoch": 3.8051118210862622, + "grad_norm": 0.435546875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 4764 + }, + { + "epoch": 3.8059105431309903, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 4765 + }, + { + "epoch": 3.806709265175719, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 4766 + }, + { + "epoch": 3.8075079872204474, + "grad_norm": 0.396484375, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 4767 + }, + { + "epoch": 3.8083067092651754, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 4768 + }, + { + "epoch": 3.809105431309904, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1283, + "step": 4769 + }, + { + "epoch": 3.8099041533546325, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 4770 + }, + { + "epoch": 3.810702875399361, + "grad_norm": 0.390625, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 4771 + }, + { + "epoch": 3.8115015974440896, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 4772 + }, + { + "epoch": 3.812300319488818, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 4773 + }, + { + "epoch": 3.813099041533546, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 4774 + }, + { + "epoch": 3.8138977635782747, + "grad_norm": 0.53125, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 4775 + }, + { + "epoch": 3.8146964856230032, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 4776 + }, + { + "epoch": 3.8154952076677318, + "grad_norm": 0.5625, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 4777 + }, + { + "epoch": 3.81629392971246, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 4778 + }, + { + "epoch": 3.8170926517571884, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 4779 + }, + { + "epoch": 3.817891373801917, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 4780 + }, + { + "epoch": 3.8186900958466454, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 4781 + }, + { + "epoch": 3.819488817891374, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 4782 + }, + { + "epoch": 3.8202875399361025, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 4783 + }, + { + "epoch": 3.8210862619808306, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 4784 + }, + { + "epoch": 3.821884984025559, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 4785 + }, + { + "epoch": 3.8226837060702876, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 4786 + }, + { + "epoch": 3.8234824281150157, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 4787 + }, + { + "epoch": 3.8242811501597442, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 4788 + }, + { + "epoch": 3.8250798722044728, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 4789 + }, + { + "epoch": 3.8258785942492013, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 4790 + }, + { + "epoch": 3.82667731629393, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 4791 + }, + { + "epoch": 3.8274760383386583, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 4792 + }, + { + "epoch": 3.8282747603833864, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 4793 + }, + { + "epoch": 3.829073482428115, + "grad_norm": 0.375, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 4794 + }, + { + "epoch": 3.8298722044728435, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4795 + }, + { + "epoch": 3.830670926517572, + "grad_norm": 0.5234375, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 4796 + }, + { + "epoch": 3.8314696485623, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4797 + }, + { + "epoch": 3.8322683706070286, + "grad_norm": 0.45703125, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 4798 + }, + { + "epoch": 3.833067092651757, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 4799 + }, + { + "epoch": 3.8338658146964857, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 4800 + }, + { + "epoch": 3.834664536741214, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 4801 + }, + { + "epoch": 3.8354632587859427, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 4802 + }, + { + "epoch": 3.836261980830671, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 4803 + }, + { + "epoch": 3.8370607028753994, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 4804 + }, + { + "epoch": 3.837859424920128, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 4805 + }, + { + "epoch": 3.838658146964856, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 4806 + }, + { + "epoch": 3.8394568690095845, + "grad_norm": 0.328125, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 4807 + }, + { + "epoch": 3.840255591054313, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 4808 + }, + { + "epoch": 3.8410543130990416, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 4809 + }, + { + "epoch": 3.84185303514377, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1235, + "step": 4810 + }, + { + "epoch": 3.8426517571884986, + "grad_norm": 0.4140625, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 4811 + }, + { + "epoch": 3.8434504792332267, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 4812 + }, + { + "epoch": 3.844249201277955, + "grad_norm": 0.33984375, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 4813 + }, + { + "epoch": 3.8450479233226837, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 4814 + }, + { + "epoch": 3.8458466453674123, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 4815 + }, + { + "epoch": 3.8466453674121404, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 4816 + }, + { + "epoch": 3.847444089456869, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 4817 + }, + { + "epoch": 3.8482428115015974, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4818 + }, + { + "epoch": 3.849041533546326, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 4819 + }, + { + "epoch": 3.8498402555910545, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 4820 + }, + { + "epoch": 3.850638977635783, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1248, + "step": 4821 + }, + { + "epoch": 3.851437699680511, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4822 + }, + { + "epoch": 3.8522364217252396, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 4823 + }, + { + "epoch": 3.853035143769968, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 4824 + }, + { + "epoch": 3.8538338658146962, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 4825 + }, + { + "epoch": 3.8546325878594248, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 4826 + }, + { + "epoch": 3.8554313099041533, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 4827 + }, + { + "epoch": 3.856230031948882, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 4828 + }, + { + "epoch": 3.8570287539936103, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 4829 + }, + { + "epoch": 3.857827476038339, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 4830 + }, + { + "epoch": 3.858626198083067, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 4831 + }, + { + "epoch": 3.8594249201277955, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 4832 + }, + { + "epoch": 3.860223642172524, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 4833 + }, + { + "epoch": 3.8610223642172525, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 4834 + }, + { + "epoch": 3.8618210862619806, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 4835 + }, + { + "epoch": 3.862619808306709, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 4836 + }, + { + "epoch": 3.8634185303514377, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 4837 + }, + { + "epoch": 3.864217252396166, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 4838 + }, + { + "epoch": 3.8650159744408947, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 4839 + }, + { + "epoch": 3.8658146964856233, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 4840 + }, + { + "epoch": 3.8666134185303513, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 4841 + }, + { + "epoch": 3.86741214057508, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1231, + "step": 4842 + }, + { + "epoch": 3.8682108626198084, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 4843 + }, + { + "epoch": 3.8690095846645365, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1268, + "step": 4844 + }, + { + "epoch": 3.869808306709265, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 4845 + }, + { + "epoch": 3.8706070287539935, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 4846 + }, + { + "epoch": 3.871405750798722, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 4847 + }, + { + "epoch": 3.8722044728434506, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 4848 + }, + { + "epoch": 3.873003194888179, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 4849 + }, + { + "epoch": 3.873801916932907, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.128, + "step": 4850 + }, + { + "epoch": 3.8746006389776357, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 4851 + }, + { + "epoch": 3.8753993610223643, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 4852 + }, + { + "epoch": 3.876198083067093, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 4853 + }, + { + "epoch": 3.876996805111821, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 4854 + }, + { + "epoch": 3.8777955271565494, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 4855 + }, + { + "epoch": 3.878594249201278, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 4856 + }, + { + "epoch": 3.8793929712460065, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 4857 + }, + { + "epoch": 3.880191693290735, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 4858 + }, + { + "epoch": 3.8809904153354635, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4859 + }, + { + "epoch": 3.8817891373801916, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4860 + }, + { + "epoch": 3.88258785942492, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 4861 + }, + { + "epoch": 3.8833865814696487, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 4862 + }, + { + "epoch": 3.8841853035143767, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 4863 + }, + { + "epoch": 3.8849840255591053, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 4864 + }, + { + "epoch": 3.885782747603834, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 4865 + }, + { + "epoch": 3.8865814696485623, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 4866 + }, + { + "epoch": 3.887380191693291, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 4867 + }, + { + "epoch": 3.8881789137380194, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 4868 + }, + { + "epoch": 3.8889776357827475, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 4869 + }, + { + "epoch": 3.889776357827476, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 4870 + }, + { + "epoch": 3.8905750798722045, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 4871 + }, + { + "epoch": 3.891373801916933, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 4872 + }, + { + "epoch": 3.892172523961661, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 4873 + }, + { + "epoch": 3.8929712460063897, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 4874 + }, + { + "epoch": 3.893769968051118, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 4875 + }, + { + "epoch": 3.8945686900958467, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 4876 + }, + { + "epoch": 3.8953674121405752, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4877 + }, + { + "epoch": 3.8961661341853038, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4878 + }, + { + "epoch": 3.896964856230032, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 4879 + }, + { + "epoch": 3.8977635782747604, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4880 + }, + { + "epoch": 3.898562300319489, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 4881 + }, + { + "epoch": 3.899361022364217, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 4882 + }, + { + "epoch": 3.9001597444089455, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 4883 + }, + { + "epoch": 3.900958466453674, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4884 + }, + { + "epoch": 3.9017571884984026, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 4885 + }, + { + "epoch": 3.902555910543131, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 4886 + }, + { + "epoch": 3.9033546325878596, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 4887 + }, + { + "epoch": 3.9041533546325877, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 4888 + }, + { + "epoch": 3.9049520766773163, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 4889 + }, + { + "epoch": 3.905750798722045, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 4890 + }, + { + "epoch": 3.9065495207667733, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 4891 + }, + { + "epoch": 3.9073482428115014, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 4892 + }, + { + "epoch": 3.90814696485623, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 4893 + }, + { + "epoch": 3.9089456869009584, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 4894 + }, + { + "epoch": 3.909744408945687, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 4895 + }, + { + "epoch": 3.9105431309904155, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 4896 + }, + { + "epoch": 3.911341853035144, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 4897 + }, + { + "epoch": 3.912140575079872, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 4898 + }, + { + "epoch": 3.9129392971246006, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 4899 + }, + { + "epoch": 3.913738019169329, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 4900 + }, + { + "epoch": 3.9145367412140573, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 4901 + }, + { + "epoch": 3.915335463258786, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 4902 + }, + { + "epoch": 3.9161341853035143, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 4903 + }, + { + "epoch": 3.916932907348243, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 4904 + }, + { + "epoch": 3.9177316293929714, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 4905 + }, + { + "epoch": 3.9185303514377, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1338, + "step": 4906 + }, + { + "epoch": 3.919329073482428, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 4907 + }, + { + "epoch": 3.9201277955271565, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 4908 + }, + { + "epoch": 3.920926517571885, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 4909 + }, + { + "epoch": 3.9217252396166136, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 4910 + }, + { + "epoch": 3.9225239616613417, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1259, + "step": 4911 + }, + { + "epoch": 3.92332268370607, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 4912 + }, + { + "epoch": 3.9241214057507987, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 4913 + }, + { + "epoch": 3.9249201277955272, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 4914 + }, + { + "epoch": 3.9257188498402558, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 4915 + }, + { + "epoch": 3.9265175718849843, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1285, + "step": 4916 + }, + { + "epoch": 3.9273162939297124, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4917 + }, + { + "epoch": 3.928115015974441, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 4918 + }, + { + "epoch": 3.9289137380191694, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4919 + }, + { + "epoch": 3.9297124600638975, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 4920 + }, + { + "epoch": 3.930511182108626, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 4921 + }, + { + "epoch": 3.9313099041533546, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 4922 + }, + { + "epoch": 3.932108626198083, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 4923 + }, + { + "epoch": 3.9329073482428116, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 4924 + }, + { + "epoch": 3.93370607028754, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 4925 + }, + { + "epoch": 3.9345047923322682, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 4926 + }, + { + "epoch": 3.9353035143769968, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 4927 + }, + { + "epoch": 3.9361022364217253, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 4928 + }, + { + "epoch": 3.936900958466454, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 4929 + }, + { + "epoch": 3.937699680511182, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 4930 + }, + { + "epoch": 3.9384984025559104, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 4931 + }, + { + "epoch": 3.939297124600639, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 4932 + }, + { + "epoch": 3.9400958466453675, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 4933 + }, + { + "epoch": 3.940894568690096, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 4934 + }, + { + "epoch": 3.9416932907348246, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 4935 + }, + { + "epoch": 3.9424920127795526, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 4936 + }, + { + "epoch": 3.943290734824281, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 4937 + }, + { + "epoch": 3.9440894568690097, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 4938 + }, + { + "epoch": 3.9448881789137378, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4939 + }, + { + "epoch": 3.9456869009584663, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 4940 + }, + { + "epoch": 3.946485623003195, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 4941 + }, + { + "epoch": 3.9472843450479234, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 4942 + }, + { + "epoch": 3.948083067092652, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 4943 + }, + { + "epoch": 3.9488817891373804, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 4944 + }, + { + "epoch": 3.9496805111821085, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 4945 + }, + { + "epoch": 3.950479233226837, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 4946 + }, + { + "epoch": 3.9512779552715656, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 4947 + }, + { + "epoch": 3.952076677316294, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 4948 + }, + { + "epoch": 3.952875399361022, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 4949 + }, + { + "epoch": 3.9536741214057507, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 4950 + }, + { + "epoch": 3.9544728434504792, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 4951 + }, + { + "epoch": 3.9552715654952078, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 4952 + }, + { + "epoch": 3.9560702875399363, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 4953 + }, + { + "epoch": 3.956869009584665, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 4954 + }, + { + "epoch": 3.957667731629393, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 4955 + }, + { + "epoch": 3.9584664536741214, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 4956 + }, + { + "epoch": 3.95926517571885, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 4957 + }, + { + "epoch": 3.960063897763578, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 4958 + }, + { + "epoch": 3.9608626198083066, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 4959 + }, + { + "epoch": 3.961661341853035, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 4960 + }, + { + "epoch": 3.9624600638977636, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 4961 + }, + { + "epoch": 3.963258785942492, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 4962 + }, + { + "epoch": 3.9640575079872207, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 4963 + }, + { + "epoch": 3.9648562300319488, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 4964 + }, + { + "epoch": 3.9656549520766773, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 4965 + }, + { + "epoch": 3.966453674121406, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 4966 + }, + { + "epoch": 3.9672523961661343, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 4967 + }, + { + "epoch": 3.9680511182108624, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 4968 + }, + { + "epoch": 3.968849840255591, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1389, + "step": 4969 + }, + { + "epoch": 3.9696485623003195, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 4970 + }, + { + "epoch": 3.970447284345048, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 4971 + }, + { + "epoch": 3.9712460063897765, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 4972 + }, + { + "epoch": 3.972044728434505, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 4973 + }, + { + "epoch": 3.972843450479233, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 4974 + }, + { + "epoch": 3.9736421725239617, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 4975 + }, + { + "epoch": 3.97444089456869, + "grad_norm": 0.34765625, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 4976 + }, + { + "epoch": 3.9752396166134183, + "grad_norm": 0.466796875, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 4977 + }, + { + "epoch": 3.976038338658147, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 4978 + }, + { + "epoch": 3.9768370607028753, + "grad_norm": 0.4453125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 4979 + }, + { + "epoch": 3.977635782747604, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 4980 + }, + { + "epoch": 3.9784345047923324, + "grad_norm": 0.439453125, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 4981 + }, + { + "epoch": 3.979233226837061, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 4982 + }, + { + "epoch": 3.980031948881789, + "grad_norm": 0.376953125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 4983 + }, + { + "epoch": 3.9808306709265175, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 4984 + }, + { + "epoch": 3.981629392971246, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 4985 + }, + { + "epoch": 3.9824281150159746, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 4986 + }, + { + "epoch": 3.9832268370607027, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 4987 + }, + { + "epoch": 3.984025559105431, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 4988 + }, + { + "epoch": 3.9848242811501597, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 4989 + }, + { + "epoch": 3.9856230031948883, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 4990 + }, + { + "epoch": 3.986421725239617, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 4991 + }, + { + "epoch": 3.987220447284345, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 4992 + }, + { + "epoch": 3.9880191693290734, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 4993 + }, + { + "epoch": 3.988817891373802, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 4994 + }, + { + "epoch": 3.9896166134185305, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 4995 + }, + { + "epoch": 3.9904153354632586, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 4996 + }, + { + "epoch": 3.991214057507987, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1279, + "step": 4997 + }, + { + "epoch": 3.9920127795527156, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 4998 + }, + { + "epoch": 3.992811501597444, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 4999 + }, + { + "epoch": 3.9936102236421727, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 5000 + }, + { + "epoch": 3.994408945686901, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 5001 + }, + { + "epoch": 3.9952076677316293, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 5002 + }, + { + "epoch": 3.996006389776358, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 5003 + }, + { + "epoch": 3.9968051118210863, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 5004 + }, + { + "epoch": 3.997603833865815, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 5005 + }, + { + "epoch": 3.998402555910543, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 5006 + }, + { + "epoch": 3.9992012779552715, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 5007 + }, + { + "epoch": 4.0, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 5008 + }, + { + "epoch": 4.0007987220447285, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5009 + }, + { + "epoch": 4.001597444089457, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 5010 + }, + { + "epoch": 4.002396166134186, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 5011 + }, + { + "epoch": 4.003194888178914, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 5012 + }, + { + "epoch": 4.003993610223642, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 5013 + }, + { + "epoch": 4.00479233226837, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 5014 + }, + { + "epoch": 4.005591054313099, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 5015 + }, + { + "epoch": 4.006389776357827, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 5016 + }, + { + "epoch": 4.007188498402556, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 5017 + }, + { + "epoch": 4.007987220447284, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 5018 + }, + { + "epoch": 4.008785942492013, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 5019 + }, + { + "epoch": 4.0095846645367414, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 5020 + }, + { + "epoch": 4.01038338658147, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 5021 + }, + { + "epoch": 4.0111821086261985, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 5022 + }, + { + "epoch": 4.011980830670926, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 5023 + }, + { + "epoch": 4.012779552715655, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 5024 + }, + { + "epoch": 4.013578274760383, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 5025 + }, + { + "epoch": 4.014376996805112, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 5026 + }, + { + "epoch": 4.01517571884984, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 5027 + }, + { + "epoch": 4.015974440894569, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 5028 + }, + { + "epoch": 4.016773162939297, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 5029 + }, + { + "epoch": 4.017571884984026, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 5030 + }, + { + "epoch": 4.018370607028754, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 5031 + }, + { + "epoch": 4.019169329073482, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 5032 + }, + { + "epoch": 4.0199680511182105, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 5033 + }, + { + "epoch": 4.020766773162939, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 5034 + }, + { + "epoch": 4.021565495207668, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 5035 + }, + { + "epoch": 4.022364217252396, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 5036 + }, + { + "epoch": 4.023162939297125, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 5037 + }, + { + "epoch": 4.023961661341853, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 5038 + }, + { + "epoch": 4.024760383386582, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 5039 + }, + { + "epoch": 4.02555910543131, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 5040 + }, + { + "epoch": 4.026357827476039, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 5041 + }, + { + "epoch": 4.027156549520766, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 5042 + }, + { + "epoch": 4.027955271565495, + "grad_norm": 0.396484375, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 5043 + }, + { + "epoch": 4.0287539936102235, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 5044 + }, + { + "epoch": 4.029552715654952, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 5045 + }, + { + "epoch": 4.0303514376996805, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 5046 + }, + { + "epoch": 4.031150159744409, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 5047 + }, + { + "epoch": 4.031948881789138, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 5048 + }, + { + "epoch": 4.032747603833866, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 5049 + }, + { + "epoch": 4.033546325878595, + "grad_norm": 0.40625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 5050 + }, + { + "epoch": 4.034345047923322, + "grad_norm": 0.5078125, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 5051 + }, + { + "epoch": 4.035143769968051, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 5052 + }, + { + "epoch": 4.035942492012779, + "grad_norm": 0.361328125, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 5053 + }, + { + "epoch": 4.036741214057508, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 5054 + }, + { + "epoch": 4.037539936102236, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 5055 + }, + { + "epoch": 4.038338658146965, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 5056 + }, + { + "epoch": 4.039137380191693, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 5057 + }, + { + "epoch": 4.039936102236422, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 5058 + }, + { + "epoch": 4.0407348242811505, + "grad_norm": 0.4453125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 5059 + }, + { + "epoch": 4.041533546325879, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 5060 + }, + { + "epoch": 4.042332268370607, + "grad_norm": 0.578125, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 5061 + }, + { + "epoch": 4.043130990415335, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 5062 + }, + { + "epoch": 4.043929712460064, + "grad_norm": 0.5078125, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 5063 + }, + { + "epoch": 4.044728434504792, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 5064 + }, + { + "epoch": 4.045527156549521, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 5065 + }, + { + "epoch": 4.046325878594249, + "grad_norm": 0.40625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 5066 + }, + { + "epoch": 4.047124600638978, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 5067 + }, + { + "epoch": 4.047923322683706, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 5068 + }, + { + "epoch": 4.048722044728435, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 5069 + }, + { + "epoch": 4.0495207667731625, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 5070 + }, + { + "epoch": 4.050319488817891, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 5071 + }, + { + "epoch": 4.05111821086262, + "grad_norm": 0.439453125, + "learning_rate": 0.0001, + "loss": 1.1328, + "step": 5072 + }, + { + "epoch": 4.051916932907348, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 5073 + }, + { + "epoch": 4.052715654952077, + "grad_norm": 0.4375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 5074 + }, + { + "epoch": 4.053514376996805, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 5075 + }, + { + "epoch": 4.054313099041534, + "grad_norm": 0.68359375, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 5076 + }, + { + "epoch": 4.055111821086262, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 5077 + }, + { + "epoch": 4.055910543130991, + "grad_norm": 0.56640625, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 5078 + }, + { + "epoch": 4.056709265175719, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 5079 + }, + { + "epoch": 4.057507987220447, + "grad_norm": 0.33984375, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 5080 + }, + { + "epoch": 4.0583067092651754, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 5081 + }, + { + "epoch": 4.059105431309904, + "grad_norm": 0.4296875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5082 + }, + { + "epoch": 4.0599041533546325, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 5083 + }, + { + "epoch": 4.060702875399361, + "grad_norm": 0.48046875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 5084 + }, + { + "epoch": 4.06150159744409, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5085 + }, + { + "epoch": 4.062300319488818, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 5086 + }, + { + "epoch": 4.063099041533547, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 5087 + }, + { + "epoch": 4.063897763578275, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 5088 + }, + { + "epoch": 4.064696485623003, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 5089 + }, + { + "epoch": 4.065495207667731, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 5090 + }, + { + "epoch": 4.06629392971246, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 5091 + }, + { + "epoch": 4.067092651757188, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 5092 + }, + { + "epoch": 4.067891373801917, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 5093 + }, + { + "epoch": 4.068690095846645, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 5094 + }, + { + "epoch": 4.069488817891374, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 5095 + }, + { + "epoch": 4.0702875399361025, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 5096 + }, + { + "epoch": 4.071086261980831, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 5097 + }, + { + "epoch": 4.0718849840255595, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.132, + "step": 5098 + }, + { + "epoch": 4.072683706070287, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 5099 + }, + { + "epoch": 4.073482428115016, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 5100 + }, + { + "epoch": 4.074281150159744, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5101 + }, + { + "epoch": 4.075079872204473, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 5102 + }, + { + "epoch": 4.075878594249201, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 5103 + }, + { + "epoch": 4.07667731629393, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 5104 + }, + { + "epoch": 4.077476038338658, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 5105 + }, + { + "epoch": 4.078274760383387, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 5106 + }, + { + "epoch": 4.079073482428115, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 5107 + }, + { + "epoch": 4.079872204472843, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 5108 + }, + { + "epoch": 4.080670926517572, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 5109 + }, + { + "epoch": 4.0814696485623, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 5110 + }, + { + "epoch": 4.082268370607029, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 5111 + }, + { + "epoch": 4.083067092651757, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 5112 + }, + { + "epoch": 4.083865814696486, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 5113 + }, + { + "epoch": 4.084664536741214, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 5114 + }, + { + "epoch": 4.085463258785943, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 5115 + }, + { + "epoch": 4.086261980830671, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 5116 + }, + { + "epoch": 4.0870607028754, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5117 + }, + { + "epoch": 4.087859424920127, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 5118 + }, + { + "epoch": 4.088658146964856, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 5119 + }, + { + "epoch": 4.0894568690095845, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5120 + }, + { + "epoch": 4.090255591054313, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 5121 + }, + { + "epoch": 4.0910543130990416, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 5122 + }, + { + "epoch": 4.09185303514377, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 5123 + }, + { + "epoch": 4.092651757188499, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 5124 + }, + { + "epoch": 4.093450479233227, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 5125 + }, + { + "epoch": 4.094249201277956, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 5126 + }, + { + "epoch": 4.095047923322683, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 5127 + }, + { + "epoch": 4.095846645367412, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 5128 + }, + { + "epoch": 4.09664536741214, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 5129 + }, + { + "epoch": 4.097444089456869, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 5130 + }, + { + "epoch": 4.098242811501597, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 5131 + }, + { + "epoch": 4.099041533546326, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 5132 + }, + { + "epoch": 4.0998402555910545, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 5133 + }, + { + "epoch": 4.100638977635783, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 5134 + }, + { + "epoch": 4.1014376996805115, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 5135 + }, + { + "epoch": 4.102236421725239, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 5136 + }, + { + "epoch": 4.103035143769968, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 5137 + }, + { + "epoch": 4.103833865814696, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 5138 + }, + { + "epoch": 4.104632587859425, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 5139 + }, + { + "epoch": 4.105431309904153, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 5140 + }, + { + "epoch": 4.106230031948882, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 5141 + }, + { + "epoch": 4.10702875399361, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 5142 + }, + { + "epoch": 4.107827476038339, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 5143 + }, + { + "epoch": 4.108626198083067, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 5144 + }, + { + "epoch": 4.109424920127796, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 5145 + }, + { + "epoch": 4.110223642172524, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 5146 + }, + { + "epoch": 4.111022364217252, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 5147 + }, + { + "epoch": 4.111821086261981, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 5148 + }, + { + "epoch": 4.112619808306709, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 5149 + }, + { + "epoch": 4.113418530351438, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 5150 + }, + { + "epoch": 4.114217252396166, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 5151 + }, + { + "epoch": 4.115015974440895, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 5152 + }, + { + "epoch": 4.115814696485623, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 5153 + }, + { + "epoch": 4.116613418530352, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 5154 + }, + { + "epoch": 4.11741214057508, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 5155 + }, + { + "epoch": 4.118210862619808, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 5156 + }, + { + "epoch": 4.1190095846645365, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5157 + }, + { + "epoch": 4.119808306709265, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1243, + "step": 5158 + }, + { + "epoch": 4.1206070287539935, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 5159 + }, + { + "epoch": 4.121405750798722, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 5160 + }, + { + "epoch": 4.122204472843451, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 5161 + }, + { + "epoch": 4.123003194888179, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 5162 + }, + { + "epoch": 4.123801916932908, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.123, + "step": 5163 + }, + { + "epoch": 4.124600638977636, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 5164 + }, + { + "epoch": 4.125399361022364, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 5165 + }, + { + "epoch": 4.126198083067092, + "grad_norm": 0.40234375, + "learning_rate": 0.0001, + "loss": 1.1306, + "step": 5166 + }, + { + "epoch": 4.126996805111821, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 5167 + }, + { + "epoch": 4.127795527156549, + "grad_norm": 0.44921875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 5168 + }, + { + "epoch": 4.128594249201278, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 5169 + }, + { + "epoch": 4.1293929712460065, + "grad_norm": 0.412109375, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 5170 + }, + { + "epoch": 4.130191693290735, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 5171 + }, + { + "epoch": 4.1309904153354635, + "grad_norm": 0.44921875, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 5172 + }, + { + "epoch": 4.131789137380192, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 5173 + }, + { + "epoch": 4.13258785942492, + "grad_norm": 0.5703125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 5174 + }, + { + "epoch": 4.133386581469648, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 5175 + }, + { + "epoch": 4.134185303514377, + "grad_norm": 0.5546875, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 5176 + }, + { + "epoch": 4.134984025559105, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 5177 + }, + { + "epoch": 4.135782747603834, + "grad_norm": 0.5625, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 5178 + }, + { + "epoch": 4.136581469648562, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 5179 + }, + { + "epoch": 4.137380191693291, + "grad_norm": 0.380859375, + "learning_rate": 0.0001, + "loss": 1.1298, + "step": 5180 + }, + { + "epoch": 4.138178913738019, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 5181 + }, + { + "epoch": 4.138977635782748, + "grad_norm": 0.484375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 5182 + }, + { + "epoch": 4.139776357827476, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 5183 + }, + { + "epoch": 4.140575079872204, + "grad_norm": 0.345703125, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 5184 + }, + { + "epoch": 4.141373801916933, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 5185 + }, + { + "epoch": 4.142172523961661, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 5186 + }, + { + "epoch": 4.14297124600639, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 5187 + }, + { + "epoch": 4.143769968051118, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 5188 + }, + { + "epoch": 4.144568690095847, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 5189 + }, + { + "epoch": 4.145367412140575, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 5190 + }, + { + "epoch": 4.146166134185304, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 5191 + }, + { + "epoch": 4.146964856230032, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 5192 + }, + { + "epoch": 4.147763578274761, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 5193 + }, + { + "epoch": 4.1485623003194885, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 5194 + }, + { + "epoch": 4.149361022364217, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 5195 + }, + { + "epoch": 4.1501597444089455, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 5196 + }, + { + "epoch": 4.150958466453674, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 5197 + }, + { + "epoch": 4.151757188498403, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 5198 + }, + { + "epoch": 4.152555910543131, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 5199 + }, + { + "epoch": 4.15335463258786, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 5200 + }, + { + "epoch": 4.154153354632588, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 5201 + }, + { + "epoch": 4.154952076677317, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 5202 + }, + { + "epoch": 4.155750798722044, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 5203 + }, + { + "epoch": 4.156549520766773, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 5204 + }, + { + "epoch": 4.157348242811501, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 5205 + }, + { + "epoch": 4.15814696485623, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 5206 + }, + { + "epoch": 4.1589456869009584, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 5207 + }, + { + "epoch": 4.159744408945687, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 5208 + }, + { + "epoch": 4.1605431309904155, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 5209 + }, + { + "epoch": 4.161341853035144, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 5210 + }, + { + "epoch": 4.162140575079873, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 5211 + }, + { + "epoch": 4.1629392971246, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 5212 + }, + { + "epoch": 4.163738019169329, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 5213 + }, + { + "epoch": 4.164536741214057, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 5214 + }, + { + "epoch": 4.165335463258786, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 5215 + }, + { + "epoch": 4.166134185303514, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 5216 + }, + { + "epoch": 4.166932907348243, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 5217 + }, + { + "epoch": 4.167731629392971, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 5218 + }, + { + "epoch": 4.1685303514377, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 5219 + }, + { + "epoch": 4.169329073482428, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 5220 + }, + { + "epoch": 4.170127795527157, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5221 + }, + { + "epoch": 4.170926517571885, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 5222 + }, + { + "epoch": 4.171725239616613, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 5223 + }, + { + "epoch": 4.172523961661342, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 5224 + }, + { + "epoch": 4.17332268370607, + "grad_norm": 0.44921875, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 5225 + }, + { + "epoch": 4.174121405750799, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 5226 + }, + { + "epoch": 4.174920127795527, + "grad_norm": 0.4921875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 5227 + }, + { + "epoch": 4.175718849840256, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5228 + }, + { + "epoch": 4.176517571884984, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 5229 + }, + { + "epoch": 4.177316293929713, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 5230 + }, + { + "epoch": 4.178115015974441, + "grad_norm": 0.423828125, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 5231 + }, + { + "epoch": 4.178913738019169, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 5232 + }, + { + "epoch": 4.1797124600638975, + "grad_norm": 0.51953125, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 5233 + }, + { + "epoch": 4.180511182108626, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 5234 + }, + { + "epoch": 4.181309904153355, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 5235 + }, + { + "epoch": 4.182108626198083, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 5236 + }, + { + "epoch": 4.182907348242812, + "grad_norm": 0.361328125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 5237 + }, + { + "epoch": 4.18370607028754, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5238 + }, + { + "epoch": 4.184504792332269, + "grad_norm": 0.4296875, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 5239 + }, + { + "epoch": 4.185303514376997, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 5240 + }, + { + "epoch": 4.186102236421725, + "grad_norm": 0.58984375, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 5241 + }, + { + "epoch": 4.186900958466453, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 5242 + }, + { + "epoch": 4.187699680511182, + "grad_norm": 0.435546875, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 5243 + }, + { + "epoch": 4.18849840255591, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 5244 + }, + { + "epoch": 4.189297124600639, + "grad_norm": 0.34375, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 5245 + }, + { + "epoch": 4.1900958466453675, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 5246 + }, + { + "epoch": 4.190894568690096, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 5247 + }, + { + "epoch": 4.1916932907348246, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 5248 + }, + { + "epoch": 4.192492012779553, + "grad_norm": 0.388671875, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 5249 + }, + { + "epoch": 4.193290734824281, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 5250 + }, + { + "epoch": 4.194089456869009, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.126, + "step": 5251 + }, + { + "epoch": 4.194888178913738, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 5252 + }, + { + "epoch": 4.195686900958466, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 5253 + }, + { + "epoch": 4.196485623003195, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 5254 + }, + { + "epoch": 4.197284345047923, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 5255 + }, + { + "epoch": 4.198083067092652, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 5256 + }, + { + "epoch": 4.19888178913738, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 5257 + }, + { + "epoch": 4.199680511182109, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 5258 + }, + { + "epoch": 4.2004792332268375, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 5259 + }, + { + "epoch": 4.201277955271565, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 5260 + }, + { + "epoch": 4.202076677316294, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 5261 + }, + { + "epoch": 4.202875399361022, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 5262 + }, + { + "epoch": 4.203674121405751, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 5263 + }, + { + "epoch": 4.204472843450479, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 5264 + }, + { + "epoch": 4.205271565495208, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 5265 + }, + { + "epoch": 4.206070287539936, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 5266 + }, + { + "epoch": 4.206869009584665, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 5267 + }, + { + "epoch": 4.207667731629393, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 5268 + }, + { + "epoch": 4.208466453674121, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 5269 + }, + { + "epoch": 4.2092651757188495, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 5270 + }, + { + "epoch": 4.210063897763578, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 5271 + }, + { + "epoch": 4.210862619808307, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 5272 + }, + { + "epoch": 4.211661341853035, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 5273 + }, + { + "epoch": 4.212460063897764, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 5274 + }, + { + "epoch": 4.213258785942492, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 5275 + }, + { + "epoch": 4.214057507987221, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 5276 + }, + { + "epoch": 4.214856230031949, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 5277 + }, + { + "epoch": 4.215654952076678, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 5278 + }, + { + "epoch": 4.216453674121405, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 5279 + }, + { + "epoch": 4.217252396166134, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 5280 + }, + { + "epoch": 4.218051118210862, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 5281 + }, + { + "epoch": 4.218849840255591, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 5282 + }, + { + "epoch": 4.2196485623003195, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 5283 + }, + { + "epoch": 4.220447284345048, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 5284 + }, + { + "epoch": 4.2212460063897765, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 5285 + }, + { + "epoch": 4.222044728434505, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 5286 + }, + { + "epoch": 4.222843450479234, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 5287 + }, + { + "epoch": 4.223642172523961, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 5288 + }, + { + "epoch": 4.22444089456869, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 5289 + }, + { + "epoch": 4.225239616613418, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 5290 + }, + { + "epoch": 4.226038338658147, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 5291 + }, + { + "epoch": 4.226837060702875, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 5292 + }, + { + "epoch": 4.227635782747604, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 5293 + }, + { + "epoch": 4.228434504792332, + "grad_norm": 0.373046875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 5294 + }, + { + "epoch": 4.229233226837061, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 5295 + }, + { + "epoch": 4.2300319488817895, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 5296 + }, + { + "epoch": 4.230830670926518, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 5297 + }, + { + "epoch": 4.231629392971246, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 5298 + }, + { + "epoch": 4.232428115015974, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 5299 + }, + { + "epoch": 4.233226837060703, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 5300 + }, + { + "epoch": 4.234025559105431, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 5301 + }, + { + "epoch": 4.23482428115016, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1238, + "step": 5302 + }, + { + "epoch": 4.235623003194888, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.125, + "step": 5303 + }, + { + "epoch": 4.236421725239617, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 5304 + }, + { + "epoch": 4.237220447284345, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1316, + "step": 5305 + }, + { + "epoch": 4.238019169329074, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 5306 + }, + { + "epoch": 4.2388178913738015, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 5307 + }, + { + "epoch": 4.23961661341853, + "grad_norm": 0.369140625, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 5308 + }, + { + "epoch": 4.2404153354632586, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 5309 + }, + { + "epoch": 4.241214057507987, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 5310 + }, + { + "epoch": 4.242012779552716, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1255, + "step": 5311 + }, + { + "epoch": 4.242811501597444, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 5312 + }, + { + "epoch": 4.243610223642173, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 5313 + }, + { + "epoch": 4.244408945686901, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 5314 + }, + { + "epoch": 4.24520766773163, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 5315 + }, + { + "epoch": 4.246006389776358, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1269, + "step": 5316 + }, + { + "epoch": 4.246805111821086, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 5317 + }, + { + "epoch": 4.247603833865814, + "grad_norm": 0.33984375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5318 + }, + { + "epoch": 4.248402555910543, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 5319 + }, + { + "epoch": 4.2492012779552715, + "grad_norm": 0.388671875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 5320 + }, + { + "epoch": 4.25, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 5321 + }, + { + "epoch": 4.2507987220447285, + "grad_norm": 0.435546875, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 5322 + }, + { + "epoch": 4.251597444089457, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 5323 + }, + { + "epoch": 4.252396166134186, + "grad_norm": 0.5859375, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 5324 + }, + { + "epoch": 4.253194888178914, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 5325 + }, + { + "epoch": 4.253993610223642, + "grad_norm": 0.734375, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 5326 + }, + { + "epoch": 4.25479233226837, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 5327 + }, + { + "epoch": 4.255591054313099, + "grad_norm": 0.328125, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 5328 + }, + { + "epoch": 4.256389776357827, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 5329 + }, + { + "epoch": 4.257188498402556, + "grad_norm": 0.369140625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5330 + }, + { + "epoch": 4.257987220447284, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 5331 + }, + { + "epoch": 4.258785942492013, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 5332 + }, + { + "epoch": 4.2595846645367414, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 5333 + }, + { + "epoch": 4.26038338658147, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 5334 + }, + { + "epoch": 4.261182108626198, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 5335 + }, + { + "epoch": 4.261980830670926, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 5336 + }, + { + "epoch": 4.262779552715655, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 5337 + }, + { + "epoch": 4.263578274760383, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 5338 + }, + { + "epoch": 4.264376996805112, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 5339 + }, + { + "epoch": 4.26517571884984, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 5340 + }, + { + "epoch": 4.265974440894569, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 5341 + }, + { + "epoch": 4.266773162939297, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 5342 + }, + { + "epoch": 4.267571884984026, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 5343 + }, + { + "epoch": 4.268370607028754, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 5344 + }, + { + "epoch": 4.269169329073483, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1271, + "step": 5345 + }, + { + "epoch": 4.2699680511182105, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 5346 + }, + { + "epoch": 4.270766773162939, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1322, + "step": 5347 + }, + { + "epoch": 4.271565495207668, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 5348 + }, + { + "epoch": 4.272364217252396, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 5349 + }, + { + "epoch": 4.273162939297125, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 5350 + }, + { + "epoch": 4.273961661341853, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 5351 + }, + { + "epoch": 4.274760383386582, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 5352 + }, + { + "epoch": 4.27555910543131, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 5353 + }, + { + "epoch": 4.276357827476039, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 5354 + }, + { + "epoch": 4.277156549520766, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 5355 + }, + { + "epoch": 4.277955271565495, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 5356 + }, + { + "epoch": 4.2787539936102235, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 5357 + }, + { + "epoch": 4.279552715654952, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 5358 + }, + { + "epoch": 4.2803514376996805, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 5359 + }, + { + "epoch": 4.281150159744409, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 5360 + }, + { + "epoch": 4.281948881789138, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 5361 + }, + { + "epoch": 4.282747603833866, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 5362 + }, + { + "epoch": 4.283546325878595, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 5363 + }, + { + "epoch": 4.284345047923322, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 5364 + }, + { + "epoch": 4.285143769968051, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1275, + "step": 5365 + }, + { + "epoch": 4.285942492012779, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 5366 + }, + { + "epoch": 4.286741214057508, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 5367 + }, + { + "epoch": 4.287539936102236, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5368 + }, + { + "epoch": 4.288338658146965, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 5369 + }, + { + "epoch": 4.289137380191693, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 5370 + }, + { + "epoch": 4.289936102236422, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 5371 + }, + { + "epoch": 4.2907348242811505, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 5372 + }, + { + "epoch": 4.291533546325878, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 5373 + }, + { + "epoch": 4.292332268370607, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 5374 + }, + { + "epoch": 4.293130990415335, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 5375 + }, + { + "epoch": 4.293929712460064, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 5376 + }, + { + "epoch": 4.294728434504792, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 5377 + }, + { + "epoch": 4.295527156549521, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1269, + "step": 5378 + }, + { + "epoch": 4.296325878594249, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 5379 + }, + { + "epoch": 4.297124600638978, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 5380 + }, + { + "epoch": 4.297923322683706, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 5381 + }, + { + "epoch": 4.298722044728435, + "grad_norm": 0.400390625, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 5382 + }, + { + "epoch": 4.2995207667731625, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1257, + "step": 5383 + }, + { + "epoch": 4.300319488817891, + "grad_norm": 0.36328125, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 5384 + }, + { + "epoch": 4.30111821086262, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 5385 + }, + { + "epoch": 4.301916932907348, + "grad_norm": 0.42578125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 5386 + }, + { + "epoch": 4.302715654952077, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 5387 + }, + { + "epoch": 4.303514376996805, + "grad_norm": 0.6640625, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 5388 + }, + { + "epoch": 4.304313099041534, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 5389 + }, + { + "epoch": 4.305111821086262, + "grad_norm": 0.74609375, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 5390 + }, + { + "epoch": 4.305910543130991, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 5391 + }, + { + "epoch": 4.306709265175719, + "grad_norm": 0.49609375, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 5392 + }, + { + "epoch": 4.307507987220447, + "grad_norm": 0.53125, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 5393 + }, + { + "epoch": 4.3083067092651754, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 5394 + }, + { + "epoch": 4.309105431309904, + "grad_norm": 0.7890625, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 5395 + }, + { + "epoch": 4.3099041533546325, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5396 + }, + { + "epoch": 4.310702875399361, + "grad_norm": 0.51953125, + "learning_rate": 0.0001, + "loss": 1.1312, + "step": 5397 + }, + { + "epoch": 4.31150159744409, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 5398 + }, + { + "epoch": 4.312300319488818, + "grad_norm": 0.36328125, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 5399 + }, + { + "epoch": 4.313099041533547, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 5400 + }, + { + "epoch": 4.313897763578275, + "grad_norm": 0.33984375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 5401 + }, + { + "epoch": 4.314696485623003, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 5402 + }, + { + "epoch": 4.315495207667731, + "grad_norm": 0.337890625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 5403 + }, + { + "epoch": 4.31629392971246, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 5404 + }, + { + "epoch": 4.317092651757188, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 5405 + }, + { + "epoch": 4.317891373801917, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 5406 + }, + { + "epoch": 4.318690095846645, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 5407 + }, + { + "epoch": 4.319488817891374, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 5408 + }, + { + "epoch": 4.3202875399361025, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 5409 + }, + { + "epoch": 4.321086261980831, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 5410 + }, + { + "epoch": 4.321884984025559, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 5411 + }, + { + "epoch": 4.322683706070287, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 5412 + }, + { + "epoch": 4.323482428115016, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5413 + }, + { + "epoch": 4.324281150159744, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 5414 + }, + { + "epoch": 4.325079872204473, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 5415 + }, + { + "epoch": 4.325878594249201, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 5416 + }, + { + "epoch": 4.32667731629393, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 5417 + }, + { + "epoch": 4.327476038338658, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 5418 + }, + { + "epoch": 4.328274760383387, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 5419 + }, + { + "epoch": 4.329073482428115, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 5420 + }, + { + "epoch": 4.329872204472843, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 5421 + }, + { + "epoch": 4.330670926517572, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 5422 + }, + { + "epoch": 4.3314696485623, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 5423 + }, + { + "epoch": 4.332268370607029, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 5424 + }, + { + "epoch": 4.333067092651757, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 5425 + }, + { + "epoch": 4.333865814696486, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5426 + }, + { + "epoch": 4.334664536741214, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1278, + "step": 5427 + }, + { + "epoch": 4.335463258785943, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5428 + }, + { + "epoch": 4.336261980830671, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 5429 + }, + { + "epoch": 4.3370607028754, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5430 + }, + { + "epoch": 4.337859424920127, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5431 + }, + { + "epoch": 4.338658146964856, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 5432 + }, + { + "epoch": 4.3394568690095845, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5433 + }, + { + "epoch": 4.340255591054313, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 5434 + }, + { + "epoch": 4.3410543130990416, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 5435 + }, + { + "epoch": 4.34185303514377, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 5436 + }, + { + "epoch": 4.342651757188499, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 5437 + }, + { + "epoch": 4.343450479233227, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 5438 + }, + { + "epoch": 4.344249201277956, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 5439 + }, + { + "epoch": 4.345047923322683, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 5440 + }, + { + "epoch": 4.345846645367412, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 5441 + }, + { + "epoch": 4.34664536741214, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 5442 + }, + { + "epoch": 4.347444089456869, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 5443 + }, + { + "epoch": 4.348242811501597, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 5444 + }, + { + "epoch": 4.349041533546326, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5445 + }, + { + "epoch": 4.3498402555910545, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 5446 + }, + { + "epoch": 4.350638977635783, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 5447 + }, + { + "epoch": 4.3514376996805115, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 5448 + }, + { + "epoch": 4.352236421725239, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 5449 + }, + { + "epoch": 4.353035143769968, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 5450 + }, + { + "epoch": 4.353833865814696, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 5451 + }, + { + "epoch": 4.354632587859425, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 5452 + }, + { + "epoch": 4.355431309904153, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 5453 + }, + { + "epoch": 4.356230031948882, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 5454 + }, + { + "epoch": 4.35702875399361, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 5455 + }, + { + "epoch": 4.357827476038339, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5456 + }, + { + "epoch": 4.358626198083067, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1283, + "step": 5457 + }, + { + "epoch": 4.359424920127796, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 5458 + }, + { + "epoch": 4.360223642172524, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 5459 + }, + { + "epoch": 4.361022364217252, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1297, + "step": 5460 + }, + { + "epoch": 4.361821086261981, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 5461 + }, + { + "epoch": 4.362619808306709, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 5462 + }, + { + "epoch": 4.363418530351438, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 5463 + }, + { + "epoch": 4.364217252396166, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5464 + }, + { + "epoch": 4.365015974440895, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 5465 + }, + { + "epoch": 4.365814696485623, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 5466 + }, + { + "epoch": 4.366613418530352, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 5467 + }, + { + "epoch": 4.36741214057508, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 5468 + }, + { + "epoch": 4.368210862619808, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 5469 + }, + { + "epoch": 4.3690095846645365, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 5470 + }, + { + "epoch": 4.369808306709265, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 5471 + }, + { + "epoch": 4.3706070287539935, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 5472 + }, + { + "epoch": 4.371405750798722, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 5473 + }, + { + "epoch": 4.372204472843451, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 5474 + }, + { + "epoch": 4.373003194888179, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5475 + }, + { + "epoch": 4.373801916932908, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1251, + "step": 5476 + }, + { + "epoch": 4.374600638977636, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 5477 + }, + { + "epoch": 4.375399361022364, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 5478 + }, + { + "epoch": 4.376198083067092, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 5479 + }, + { + "epoch": 4.376996805111821, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 5480 + }, + { + "epoch": 4.377795527156549, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 5481 + }, + { + "epoch": 4.378594249201278, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 5482 + }, + { + "epoch": 4.3793929712460065, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 5483 + }, + { + "epoch": 4.380191693290735, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 5484 + }, + { + "epoch": 4.3809904153354635, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 5485 + }, + { + "epoch": 4.381789137380192, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 5486 + }, + { + "epoch": 4.38258785942492, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 5487 + }, + { + "epoch": 4.383386581469648, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 5488 + }, + { + "epoch": 4.384185303514377, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 5489 + }, + { + "epoch": 4.384984025559105, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 5490 + }, + { + "epoch": 4.385782747603834, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 5491 + }, + { + "epoch": 4.386581469648562, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 5492 + }, + { + "epoch": 4.387380191693291, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 5493 + }, + { + "epoch": 4.388178913738019, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 5494 + }, + { + "epoch": 4.388977635782748, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 5495 + }, + { + "epoch": 4.389776357827476, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 5496 + }, + { + "epoch": 4.390575079872204, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 5497 + }, + { + "epoch": 4.391373801916933, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 5498 + }, + { + "epoch": 4.392172523961661, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 5499 + }, + { + "epoch": 4.39297124600639, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 5500 + }, + { + "epoch": 4.393769968051118, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 5501 + }, + { + "epoch": 4.394568690095847, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 5502 + }, + { + "epoch": 4.395367412140575, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 5503 + }, + { + "epoch": 4.396166134185304, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 5504 + }, + { + "epoch": 4.396964856230032, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5505 + }, + { + "epoch": 4.397763578274761, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 5506 + }, + { + "epoch": 4.3985623003194885, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 5507 + }, + { + "epoch": 4.399361022364217, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 5508 + }, + { + "epoch": 4.4001597444089455, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 5509 + }, + { + "epoch": 4.400958466453674, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1208, + "step": 5510 + }, + { + "epoch": 4.401757188498403, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 5511 + }, + { + "epoch": 4.402555910543131, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 5512 + }, + { + "epoch": 4.40335463258786, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 5513 + }, + { + "epoch": 4.404153354632588, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 5514 + }, + { + "epoch": 4.404952076677317, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 5515 + }, + { + "epoch": 4.405750798722044, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 5516 + }, + { + "epoch": 4.406549520766773, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5517 + }, + { + "epoch": 4.407348242811501, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 5518 + }, + { + "epoch": 4.40814696485623, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 5519 + }, + { + "epoch": 4.4089456869009584, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 5520 + }, + { + "epoch": 4.409744408945687, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 5521 + }, + { + "epoch": 4.4105431309904155, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 5522 + }, + { + "epoch": 4.411341853035144, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 5523 + }, + { + "epoch": 4.412140575079873, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 5524 + }, + { + "epoch": 4.4129392971246, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 5525 + }, + { + "epoch": 4.413738019169329, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 5526 + }, + { + "epoch": 4.414536741214057, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 5527 + }, + { + "epoch": 4.415335463258786, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 5528 + }, + { + "epoch": 4.416134185303514, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 5529 + }, + { + "epoch": 4.416932907348243, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 5530 + }, + { + "epoch": 4.417731629392971, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 5531 + }, + { + "epoch": 4.4185303514377, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 5532 + }, + { + "epoch": 4.419329073482428, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 5533 + }, + { + "epoch": 4.420127795527157, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 5534 + }, + { + "epoch": 4.420926517571885, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 5535 + }, + { + "epoch": 4.421725239616613, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 5536 + }, + { + "epoch": 4.422523961661342, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 5537 + }, + { + "epoch": 4.42332268370607, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 5538 + }, + { + "epoch": 4.424121405750799, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 5539 + }, + { + "epoch": 4.424920127795527, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5540 + }, + { + "epoch": 4.425718849840256, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5541 + }, + { + "epoch": 4.426517571884984, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 5542 + }, + { + "epoch": 4.427316293929713, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 5543 + }, + { + "epoch": 4.428115015974441, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 5544 + }, + { + "epoch": 4.428913738019169, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 5545 + }, + { + "epoch": 4.4297124600638975, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5546 + }, + { + "epoch": 4.430511182108626, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 5547 + }, + { + "epoch": 4.431309904153355, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 5548 + }, + { + "epoch": 4.432108626198083, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5549 + }, + { + "epoch": 4.432907348242812, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 5550 + }, + { + "epoch": 4.43370607028754, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 5551 + }, + { + "epoch": 4.434504792332269, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 5552 + }, + { + "epoch": 4.435303514376997, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 5553 + }, + { + "epoch": 4.436102236421725, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 5554 + }, + { + "epoch": 4.436900958466453, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 5555 + }, + { + "epoch": 4.437699680511182, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5556 + }, + { + "epoch": 4.43849840255591, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 5557 + }, + { + "epoch": 4.439297124600639, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 5558 + }, + { + "epoch": 4.4400958466453675, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 5559 + }, + { + "epoch": 4.440894568690096, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 5560 + }, + { + "epoch": 4.4416932907348246, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 5561 + }, + { + "epoch": 4.442492012779553, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5562 + }, + { + "epoch": 4.443290734824281, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 5563 + }, + { + "epoch": 4.444089456869009, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 5564 + }, + { + "epoch": 4.444888178913738, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 5565 + }, + { + "epoch": 4.445686900958466, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 5566 + }, + { + "epoch": 4.446485623003195, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 5567 + }, + { + "epoch": 4.447284345047923, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 5568 + }, + { + "epoch": 4.448083067092652, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 5569 + }, + { + "epoch": 4.44888178913738, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 5570 + }, + { + "epoch": 4.449680511182109, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 5571 + }, + { + "epoch": 4.4504792332268375, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 5572 + }, + { + "epoch": 4.451277955271565, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 5573 + }, + { + "epoch": 4.452076677316294, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 5574 + }, + { + "epoch": 4.452875399361022, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 5575 + }, + { + "epoch": 4.453674121405751, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5576 + }, + { + "epoch": 4.454472843450479, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 5577 + }, + { + "epoch": 4.455271565495208, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 5578 + }, + { + "epoch": 4.456070287539936, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 5579 + }, + { + "epoch": 4.456869009584665, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 5580 + }, + { + "epoch": 4.457667731629393, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 5581 + }, + { + "epoch": 4.458466453674122, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 5582 + }, + { + "epoch": 4.4592651757188495, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 5583 + }, + { + "epoch": 4.460063897763578, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 5584 + }, + { + "epoch": 4.460862619808307, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 5585 + }, + { + "epoch": 4.461661341853035, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 5586 + }, + { + "epoch": 4.462460063897764, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 5587 + }, + { + "epoch": 4.463258785942492, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 5588 + }, + { + "epoch": 4.464057507987221, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 5589 + }, + { + "epoch": 4.464856230031949, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 5590 + }, + { + "epoch": 4.465654952076678, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5591 + }, + { + "epoch": 4.466453674121405, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 5592 + }, + { + "epoch": 4.467252396166134, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 5593 + }, + { + "epoch": 4.468051118210862, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 5594 + }, + { + "epoch": 4.468849840255591, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 5595 + }, + { + "epoch": 4.4696485623003195, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5596 + }, + { + "epoch": 4.470447284345048, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 5597 + }, + { + "epoch": 4.4712460063897765, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 5598 + }, + { + "epoch": 4.472044728434505, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5599 + }, + { + "epoch": 4.472843450479234, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 5600 + }, + { + "epoch": 4.473642172523961, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 5601 + }, + { + "epoch": 4.47444089456869, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 5602 + }, + { + "epoch": 4.475239616613418, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5603 + }, + { + "epoch": 4.476038338658147, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 5604 + }, + { + "epoch": 4.476837060702875, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 5605 + }, + { + "epoch": 4.477635782747604, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 5606 + }, + { + "epoch": 4.478434504792332, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 5607 + }, + { + "epoch": 4.479233226837061, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 5608 + }, + { + "epoch": 4.4800319488817895, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 5609 + }, + { + "epoch": 4.480830670926517, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1272, + "step": 5610 + }, + { + "epoch": 4.481629392971246, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1226, + "step": 5611 + }, + { + "epoch": 4.482428115015974, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 5612 + }, + { + "epoch": 4.483226837060703, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 5613 + }, + { + "epoch": 4.484025559105431, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 5614 + }, + { + "epoch": 4.48482428115016, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 5615 + }, + { + "epoch": 4.485623003194888, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 5616 + }, + { + "epoch": 4.486421725239617, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 5617 + }, + { + "epoch": 4.487220447284345, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 5618 + }, + { + "epoch": 4.488019169329074, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 5619 + }, + { + "epoch": 4.488817891373802, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 5620 + }, + { + "epoch": 4.48961661341853, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5621 + }, + { + "epoch": 4.4904153354632586, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1276, + "step": 5622 + }, + { + "epoch": 4.491214057507987, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 5623 + }, + { + "epoch": 4.492012779552716, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 5624 + }, + { + "epoch": 4.492811501597444, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 5625 + }, + { + "epoch": 4.493610223642173, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1273, + "step": 5626 + }, + { + "epoch": 4.494408945686901, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 5627 + }, + { + "epoch": 4.49520766773163, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 5628 + }, + { + "epoch": 4.496006389776358, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 5629 + }, + { + "epoch": 4.496805111821086, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 5630 + }, + { + "epoch": 4.497603833865814, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 5631 + }, + { + "epoch": 4.498402555910543, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 5632 + }, + { + "epoch": 4.4992012779552715, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 5633 + }, + { + "epoch": 4.5, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 5634 + }, + { + "epoch": 4.5007987220447285, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 5635 + }, + { + "epoch": 4.501597444089457, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 5636 + }, + { + "epoch": 4.502396166134186, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 5637 + }, + { + "epoch": 4.503194888178914, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 5638 + }, + { + "epoch": 4.503993610223642, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 5639 + }, + { + "epoch": 4.50479233226837, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 5640 + }, + { + "epoch": 4.505591054313099, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 5641 + }, + { + "epoch": 4.506389776357827, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 5642 + }, + { + "epoch": 4.507188498402556, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 5643 + }, + { + "epoch": 4.507987220447284, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 5644 + }, + { + "epoch": 4.508785942492013, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 5645 + }, + { + "epoch": 4.5095846645367414, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 5646 + }, + { + "epoch": 4.51038338658147, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 5647 + }, + { + "epoch": 4.511182108626198, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 5648 + }, + { + "epoch": 4.511980830670926, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 5649 + }, + { + "epoch": 4.512779552715655, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 5650 + }, + { + "epoch": 4.513578274760383, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 5651 + }, + { + "epoch": 4.514376996805112, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1229, + "step": 5652 + }, + { + "epoch": 4.51517571884984, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 5653 + }, + { + "epoch": 4.515974440894569, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 5654 + }, + { + "epoch": 4.516773162939297, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5655 + }, + { + "epoch": 4.517571884984026, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5656 + }, + { + "epoch": 4.518370607028754, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 5657 + }, + { + "epoch": 4.519169329073483, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 5658 + }, + { + "epoch": 4.5199680511182105, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 5659 + }, + { + "epoch": 4.520766773162939, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5660 + }, + { + "epoch": 4.521565495207668, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 5661 + }, + { + "epoch": 4.522364217252396, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 5662 + }, + { + "epoch": 4.523162939297125, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 5663 + }, + { + "epoch": 4.523961661341853, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 5664 + }, + { + "epoch": 4.524760383386582, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 5665 + }, + { + "epoch": 4.52555910543131, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 5666 + }, + { + "epoch": 4.526357827476039, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 5667 + }, + { + "epoch": 4.527156549520766, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 5668 + }, + { + "epoch": 4.527955271565495, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 5669 + }, + { + "epoch": 4.5287539936102235, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 5670 + }, + { + "epoch": 4.529552715654952, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 5671 + }, + { + "epoch": 4.5303514376996805, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 5672 + }, + { + "epoch": 4.531150159744409, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 5673 + }, + { + "epoch": 4.531948881789138, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 5674 + }, + { + "epoch": 4.532747603833866, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 5675 + }, + { + "epoch": 4.533546325878595, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 5676 + }, + { + "epoch": 4.534345047923322, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 5677 + }, + { + "epoch": 4.535143769968051, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 5678 + }, + { + "epoch": 4.535942492012779, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5679 + }, + { + "epoch": 4.536741214057508, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 5680 + }, + { + "epoch": 4.537539936102236, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 5681 + }, + { + "epoch": 4.538338658146965, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 5682 + }, + { + "epoch": 4.539137380191693, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 5683 + }, + { + "epoch": 4.539936102236422, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 5684 + }, + { + "epoch": 4.5407348242811505, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 5685 + }, + { + "epoch": 4.541533546325878, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 5686 + }, + { + "epoch": 4.542332268370607, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5687 + }, + { + "epoch": 4.543130990415335, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 5688 + }, + { + "epoch": 4.543929712460064, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 5689 + }, + { + "epoch": 4.544728434504792, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 5690 + }, + { + "epoch": 4.545527156549521, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1227, + "step": 5691 + }, + { + "epoch": 4.546325878594249, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 5692 + }, + { + "epoch": 4.547124600638978, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 5693 + }, + { + "epoch": 4.547923322683706, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 5694 + }, + { + "epoch": 4.548722044728435, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 5695 + }, + { + "epoch": 4.549520766773163, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 5696 + }, + { + "epoch": 4.550319488817891, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 5697 + }, + { + "epoch": 4.55111821086262, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 5698 + }, + { + "epoch": 4.551916932907348, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 5699 + }, + { + "epoch": 4.552715654952077, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 5700 + }, + { + "epoch": 4.553514376996805, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 5701 + }, + { + "epoch": 4.554313099041534, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 5702 + }, + { + "epoch": 4.555111821086262, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 5703 + }, + { + "epoch": 4.555910543130991, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 5704 + }, + { + "epoch": 4.556709265175719, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 5705 + }, + { + "epoch": 4.557507987220447, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5706 + }, + { + "epoch": 4.5583067092651754, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 5707 + }, + { + "epoch": 4.559105431309904, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 5708 + }, + { + "epoch": 4.5599041533546325, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 5709 + }, + { + "epoch": 4.560702875399361, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 5710 + }, + { + "epoch": 4.56150159744409, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 5711 + }, + { + "epoch": 4.562300319488818, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 5712 + }, + { + "epoch": 4.563099041533547, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 5713 + }, + { + "epoch": 4.563897763578275, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 5714 + }, + { + "epoch": 4.564696485623003, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 5715 + }, + { + "epoch": 4.565495207667731, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 5716 + }, + { + "epoch": 4.56629392971246, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 5717 + }, + { + "epoch": 4.567092651757188, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 5718 + }, + { + "epoch": 4.567891373801917, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 5719 + }, + { + "epoch": 4.568690095846645, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 5720 + }, + { + "epoch": 4.569488817891374, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 5721 + }, + { + "epoch": 4.5702875399361025, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 5722 + }, + { + "epoch": 4.571086261980831, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 5723 + }, + { + "epoch": 4.571884984025559, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 5724 + }, + { + "epoch": 4.572683706070287, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 5725 + }, + { + "epoch": 4.573482428115016, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 5726 + }, + { + "epoch": 4.574281150159744, + "grad_norm": 0.404296875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 5727 + }, + { + "epoch": 4.575079872204473, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 5728 + }, + { + "epoch": 4.575878594249201, + "grad_norm": 0.404296875, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 5729 + }, + { + "epoch": 4.57667731629393, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 5730 + }, + { + "epoch": 4.577476038338658, + "grad_norm": 0.490234375, + "learning_rate": 0.0001, + "loss": 1.1248, + "step": 5731 + }, + { + "epoch": 4.578274760383387, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 5732 + }, + { + "epoch": 4.5790734824281145, + "grad_norm": 0.5234375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 5733 + }, + { + "epoch": 4.579872204472844, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 5734 + }, + { + "epoch": 4.580670926517572, + "grad_norm": 0.61328125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 5735 + }, + { + "epoch": 4.5814696485623, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 5736 + }, + { + "epoch": 4.582268370607029, + "grad_norm": 0.353515625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 5737 + }, + { + "epoch": 4.583067092651757, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 5738 + }, + { + "epoch": 4.583865814696486, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 5739 + }, + { + "epoch": 4.584664536741214, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 5740 + }, + { + "epoch": 4.585463258785943, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 5741 + }, + { + "epoch": 4.586261980830671, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 5742 + }, + { + "epoch": 4.5870607028754, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 5743 + }, + { + "epoch": 4.587859424920127, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 5744 + }, + { + "epoch": 4.588658146964856, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 5745 + }, + { + "epoch": 4.5894568690095845, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 5746 + }, + { + "epoch": 4.590255591054313, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 5747 + }, + { + "epoch": 4.5910543130990416, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 5748 + }, + { + "epoch": 4.59185303514377, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 5749 + }, + { + "epoch": 4.592651757188499, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 5750 + }, + { + "epoch": 4.593450479233227, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 5751 + }, + { + "epoch": 4.594249201277956, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 5752 + }, + { + "epoch": 4.595047923322683, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 5753 + }, + { + "epoch": 4.595846645367412, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 5754 + }, + { + "epoch": 4.59664536741214, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 5755 + }, + { + "epoch": 4.597444089456869, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 5756 + }, + { + "epoch": 4.598242811501597, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 5757 + }, + { + "epoch": 4.599041533546326, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 5758 + }, + { + "epoch": 4.5998402555910545, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 5759 + }, + { + "epoch": 4.600638977635783, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 5760 + }, + { + "epoch": 4.6014376996805115, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 5761 + }, + { + "epoch": 4.602236421725239, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 5762 + }, + { + "epoch": 4.603035143769968, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 5763 + }, + { + "epoch": 4.603833865814696, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5764 + }, + { + "epoch": 4.604632587859425, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 5765 + }, + { + "epoch": 4.605431309904153, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 5766 + }, + { + "epoch": 4.606230031948882, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 5767 + }, + { + "epoch": 4.60702875399361, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 5768 + }, + { + "epoch": 4.607827476038339, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 5769 + }, + { + "epoch": 4.608626198083067, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5770 + }, + { + "epoch": 4.609424920127795, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 5771 + }, + { + "epoch": 4.6102236421725244, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 5772 + }, + { + "epoch": 4.611022364217252, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 5773 + }, + { + "epoch": 4.611821086261981, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5774 + }, + { + "epoch": 4.612619808306709, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 5775 + }, + { + "epoch": 4.613418530351438, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 5776 + }, + { + "epoch": 4.614217252396166, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 5777 + }, + { + "epoch": 4.615015974440895, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 5778 + }, + { + "epoch": 4.615814696485623, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 5779 + }, + { + "epoch": 4.616613418530352, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 5780 + }, + { + "epoch": 4.61741214057508, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 5781 + }, + { + "epoch": 4.618210862619808, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 5782 + }, + { + "epoch": 4.6190095846645365, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5783 + }, + { + "epoch": 4.619808306709265, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 5784 + }, + { + "epoch": 4.6206070287539935, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 5785 + }, + { + "epoch": 4.621405750798722, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 5786 + }, + { + "epoch": 4.622204472843451, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 5787 + }, + { + "epoch": 4.623003194888179, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 5788 + }, + { + "epoch": 4.623801916932908, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 5789 + }, + { + "epoch": 4.624600638977636, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 5790 + }, + { + "epoch": 4.625399361022364, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 5791 + }, + { + "epoch": 4.626198083067092, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 5792 + }, + { + "epoch": 4.626996805111821, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 5793 + }, + { + "epoch": 4.627795527156549, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1362, + "step": 5794 + }, + { + "epoch": 4.628594249201278, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5795 + }, + { + "epoch": 4.6293929712460065, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 5796 + }, + { + "epoch": 4.630191693290735, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 5797 + }, + { + "epoch": 4.6309904153354635, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 5798 + }, + { + "epoch": 4.631789137380192, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5799 + }, + { + "epoch": 4.63258785942492, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 5800 + }, + { + "epoch": 4.633386581469648, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 5801 + }, + { + "epoch": 4.634185303514377, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 5802 + }, + { + "epoch": 4.634984025559105, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 5803 + }, + { + "epoch": 4.635782747603834, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 5804 + }, + { + "epoch": 4.636581469648562, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 5805 + }, + { + "epoch": 4.637380191693291, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 5806 + }, + { + "epoch": 4.638178913738019, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 5807 + }, + { + "epoch": 4.638977635782748, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 5808 + }, + { + "epoch": 4.6397763578274756, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 5809 + }, + { + "epoch": 4.640575079872205, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 5810 + }, + { + "epoch": 4.641373801916933, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 5811 + }, + { + "epoch": 4.642172523961661, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 5812 + }, + { + "epoch": 4.64297124600639, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 5813 + }, + { + "epoch": 4.643769968051118, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 5814 + }, + { + "epoch": 4.644568690095847, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 5815 + }, + { + "epoch": 4.645367412140575, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 5816 + }, + { + "epoch": 4.646166134185304, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 5817 + }, + { + "epoch": 4.646964856230032, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 5818 + }, + { + "epoch": 4.647763578274761, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 5819 + }, + { + "epoch": 4.6485623003194885, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 5820 + }, + { + "epoch": 4.649361022364217, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 5821 + }, + { + "epoch": 4.6501597444089455, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 5822 + }, + { + "epoch": 4.650958466453674, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 5823 + }, + { + "epoch": 4.651757188498403, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 5824 + }, + { + "epoch": 4.652555910543131, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 5825 + }, + { + "epoch": 4.65335463258786, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 5826 + }, + { + "epoch": 4.654153354632588, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 5827 + }, + { + "epoch": 4.654952076677317, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 5828 + }, + { + "epoch": 4.655750798722044, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 5829 + }, + { + "epoch": 4.656549520766773, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 5830 + }, + { + "epoch": 4.657348242811501, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 5831 + }, + { + "epoch": 4.65814696485623, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 5832 + }, + { + "epoch": 4.6589456869009584, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 5833 + }, + { + "epoch": 4.659744408945687, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 5834 + }, + { + "epoch": 4.6605431309904155, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 5835 + }, + { + "epoch": 4.661341853035144, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 5836 + }, + { + "epoch": 4.662140575079873, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 5837 + }, + { + "epoch": 4.6629392971246, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1299, + "step": 5838 + }, + { + "epoch": 4.663738019169329, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 5839 + }, + { + "epoch": 4.664536741214057, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 5840 + }, + { + "epoch": 4.665335463258786, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 5841 + }, + { + "epoch": 4.666134185303514, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 5842 + }, + { + "epoch": 4.666932907348243, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 5843 + }, + { + "epoch": 4.667731629392971, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 5844 + }, + { + "epoch": 4.6685303514377, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 5845 + }, + { + "epoch": 4.669329073482428, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 5846 + }, + { + "epoch": 4.670127795527156, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 5847 + }, + { + "epoch": 4.6709265175718855, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 5848 + }, + { + "epoch": 4.671725239616613, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 5849 + }, + { + "epoch": 4.672523961661342, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 5850 + }, + { + "epoch": 4.67332268370607, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 5851 + }, + { + "epoch": 4.674121405750799, + "grad_norm": 0.40234375, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 5852 + }, + { + "epoch": 4.674920127795527, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 5853 + }, + { + "epoch": 4.675718849840256, + "grad_norm": 0.55078125, + "learning_rate": 0.0001, + "loss": 1.1287, + "step": 5854 + }, + { + "epoch": 4.676517571884984, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 5855 + }, + { + "epoch": 4.677316293929713, + "grad_norm": 0.4140625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 5856 + }, + { + "epoch": 4.678115015974441, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1249, + "step": 5857 + }, + { + "epoch": 4.678913738019169, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 5858 + }, + { + "epoch": 4.6797124600638975, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 5859 + }, + { + "epoch": 4.680511182108626, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 5860 + }, + { + "epoch": 4.681309904153355, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5861 + }, + { + "epoch": 4.682108626198083, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 5862 + }, + { + "epoch": 4.682907348242812, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 5863 + }, + { + "epoch": 4.68370607028754, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 5864 + }, + { + "epoch": 4.684504792332269, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 5865 + }, + { + "epoch": 4.685303514376997, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 5866 + }, + { + "epoch": 4.686102236421725, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 5867 + }, + { + "epoch": 4.686900958466453, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 5868 + }, + { + "epoch": 4.687699680511182, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 5869 + }, + { + "epoch": 4.68849840255591, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 5870 + }, + { + "epoch": 4.689297124600639, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 5871 + }, + { + "epoch": 4.6900958466453675, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 5872 + }, + { + "epoch": 4.690894568690096, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 5873 + }, + { + "epoch": 4.6916932907348246, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 5874 + }, + { + "epoch": 4.692492012779553, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 5875 + }, + { + "epoch": 4.693290734824281, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5876 + }, + { + "epoch": 4.694089456869009, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 5877 + }, + { + "epoch": 4.694888178913738, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 5878 + }, + { + "epoch": 4.695686900958466, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 5879 + }, + { + "epoch": 4.696485623003195, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5880 + }, + { + "epoch": 4.697284345047923, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 5881 + }, + { + "epoch": 4.698083067092652, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 5882 + }, + { + "epoch": 4.69888178913738, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 5883 + }, + { + "epoch": 4.699680511182109, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 5884 + }, + { + "epoch": 4.700479233226837, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 5885 + }, + { + "epoch": 4.701277955271565, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 5886 + }, + { + "epoch": 4.702076677316294, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 5887 + }, + { + "epoch": 4.702875399361022, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 5888 + }, + { + "epoch": 4.703674121405751, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 5889 + }, + { + "epoch": 4.704472843450479, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 5890 + }, + { + "epoch": 4.705271565495208, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 5891 + }, + { + "epoch": 4.706070287539936, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 5892 + }, + { + "epoch": 4.706869009584665, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 5893 + }, + { + "epoch": 4.707667731629393, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 5894 + }, + { + "epoch": 4.708466453674122, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1254, + "step": 5895 + }, + { + "epoch": 4.7092651757188495, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 5896 + }, + { + "epoch": 4.710063897763578, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 5897 + }, + { + "epoch": 4.710862619808307, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 5898 + }, + { + "epoch": 4.711661341853035, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 5899 + }, + { + "epoch": 4.712460063897764, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 5900 + }, + { + "epoch": 4.713258785942492, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 5901 + }, + { + "epoch": 4.714057507987221, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 5902 + }, + { + "epoch": 4.714856230031949, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 5903 + }, + { + "epoch": 4.715654952076678, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 5904 + }, + { + "epoch": 4.716453674121405, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 5905 + }, + { + "epoch": 4.717252396166134, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 5906 + }, + { + "epoch": 4.718051118210862, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 5907 + }, + { + "epoch": 4.718849840255591, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5908 + }, + { + "epoch": 4.7196485623003195, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 5909 + }, + { + "epoch": 4.720447284345048, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 5910 + }, + { + "epoch": 4.7212460063897765, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 5911 + }, + { + "epoch": 4.722044728434505, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 5912 + }, + { + "epoch": 4.722843450479234, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 5913 + }, + { + "epoch": 4.723642172523961, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 5914 + }, + { + "epoch": 4.72444089456869, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 5915 + }, + { + "epoch": 4.725239616613418, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 5916 + }, + { + "epoch": 4.726038338658147, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 5917 + }, + { + "epoch": 4.726837060702875, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 5918 + }, + { + "epoch": 4.727635782747604, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 5919 + }, + { + "epoch": 4.728434504792332, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 5920 + }, + { + "epoch": 4.729233226837061, + "grad_norm": 0.408203125, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5921 + }, + { + "epoch": 4.7300319488817895, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 5922 + }, + { + "epoch": 4.730830670926517, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 5923 + }, + { + "epoch": 4.731629392971246, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 5924 + }, + { + "epoch": 4.732428115015974, + "grad_norm": 0.337890625, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 5925 + }, + { + "epoch": 4.733226837060703, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 5926 + }, + { + "epoch": 4.734025559105431, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 5927 + }, + { + "epoch": 4.73482428115016, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 5928 + }, + { + "epoch": 4.735623003194888, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 5929 + }, + { + "epoch": 4.736421725239617, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 5930 + }, + { + "epoch": 4.737220447284345, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 5931 + }, + { + "epoch": 4.738019169329074, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 5932 + }, + { + "epoch": 4.738817891373802, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 5933 + }, + { + "epoch": 4.73961661341853, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 5934 + }, + { + "epoch": 4.7404153354632586, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 5935 + }, + { + "epoch": 4.741214057507987, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 5936 + }, + { + "epoch": 4.742012779552716, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 5937 + }, + { + "epoch": 4.742811501597444, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 5938 + }, + { + "epoch": 4.743610223642173, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 5939 + }, + { + "epoch": 4.744408945686901, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 5940 + }, + { + "epoch": 4.74520766773163, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 5941 + }, + { + "epoch": 4.746006389776358, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 5942 + }, + { + "epoch": 4.746805111821086, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 5943 + }, + { + "epoch": 4.747603833865814, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1284, + "step": 5944 + }, + { + "epoch": 4.748402555910543, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1277, + "step": 5945 + }, + { + "epoch": 4.7492012779552715, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 5946 + }, + { + "epoch": 4.75, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 5947 + }, + { + "epoch": 4.7507987220447285, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 5948 + }, + { + "epoch": 4.751597444089457, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 5949 + }, + { + "epoch": 4.752396166134186, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 5950 + }, + { + "epoch": 4.753194888178914, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 5951 + }, + { + "epoch": 4.753993610223642, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1247, + "step": 5952 + }, + { + "epoch": 4.75479233226837, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 5953 + }, + { + "epoch": 4.755591054313099, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 5954 + }, + { + "epoch": 4.756389776357827, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 5955 + }, + { + "epoch": 4.757188498402556, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 5956 + }, + { + "epoch": 4.757987220447284, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 5957 + }, + { + "epoch": 4.758785942492013, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1205, + "step": 5958 + }, + { + "epoch": 4.7595846645367414, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 5959 + }, + { + "epoch": 4.76038338658147, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 5960 + }, + { + "epoch": 4.761182108626198, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 5961 + }, + { + "epoch": 4.761980830670926, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 5962 + }, + { + "epoch": 4.762779552715655, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5963 + }, + { + "epoch": 4.763578274760383, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 5964 + }, + { + "epoch": 4.764376996805112, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 5965 + }, + { + "epoch": 4.76517571884984, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 5966 + }, + { + "epoch": 4.765974440894569, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 5967 + }, + { + "epoch": 4.766773162939297, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 5968 + }, + { + "epoch": 4.767571884984026, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 5969 + }, + { + "epoch": 4.768370607028754, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 5970 + }, + { + "epoch": 4.769169329073483, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 5971 + }, + { + "epoch": 4.7699680511182105, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 5972 + }, + { + "epoch": 4.770766773162939, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 5973 + }, + { + "epoch": 4.771565495207668, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 5974 + }, + { + "epoch": 4.772364217252396, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 5975 + }, + { + "epoch": 4.773162939297125, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 5976 + }, + { + "epoch": 4.773961661341853, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 5977 + }, + { + "epoch": 4.774760383386582, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 5978 + }, + { + "epoch": 4.77555910543131, + "grad_norm": 0.4296875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 5979 + }, + { + "epoch": 4.776357827476039, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1239, + "step": 5980 + }, + { + "epoch": 4.777156549520766, + "grad_norm": 0.34375, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 5981 + }, + { + "epoch": 4.777955271565495, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1234, + "step": 5982 + }, + { + "epoch": 4.7787539936102235, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 5983 + }, + { + "epoch": 4.779552715654952, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 5984 + }, + { + "epoch": 4.7803514376996805, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 5985 + }, + { + "epoch": 4.781150159744409, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 5986 + }, + { + "epoch": 4.781948881789138, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 5987 + }, + { + "epoch": 4.782747603833866, + "grad_norm": 0.388671875, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 5988 + }, + { + "epoch": 4.783546325878595, + "grad_norm": 0.412109375, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 5989 + }, + { + "epoch": 4.784345047923322, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 5990 + }, + { + "epoch": 4.785143769968051, + "grad_norm": 0.4765625, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 5991 + }, + { + "epoch": 4.785942492012779, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 5992 + }, + { + "epoch": 4.786741214057508, + "grad_norm": 0.55078125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 5993 + }, + { + "epoch": 4.787539936102236, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 5994 + }, + { + "epoch": 4.788338658146965, + "grad_norm": 0.53515625, + "learning_rate": 0.0001, + "loss": 1.1287, + "step": 5995 + }, + { + "epoch": 4.789137380191693, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 5996 + }, + { + "epoch": 4.789936102236422, + "grad_norm": 0.8046875, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 5997 + }, + { + "epoch": 4.7907348242811505, + "grad_norm": 0.453125, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 5998 + }, + { + "epoch": 4.791533546325878, + "grad_norm": 0.4375, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 5999 + }, + { + "epoch": 4.792332268370607, + "grad_norm": 0.435546875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6000 + }, + { + "epoch": 4.793130990415335, + "grad_norm": 0.390625, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 6001 + }, + { + "epoch": 4.793929712460064, + "grad_norm": 0.50390625, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 6002 + }, + { + "epoch": 4.794728434504792, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 6003 + }, + { + "epoch": 4.795527156549521, + "grad_norm": 0.37890625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 6004 + }, + { + "epoch": 4.796325878594249, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 6005 + }, + { + "epoch": 4.797124600638978, + "grad_norm": 0.337890625, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 6006 + }, + { + "epoch": 4.797923322683706, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 6007 + }, + { + "epoch": 4.798722044728435, + "grad_norm": 0.33984375, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 6008 + }, + { + "epoch": 4.799520766773163, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 6009 + }, + { + "epoch": 4.800319488817891, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 6010 + }, + { + "epoch": 4.80111821086262, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 6011 + }, + { + "epoch": 4.801916932907348, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 6012 + }, + { + "epoch": 4.802715654952077, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 6013 + }, + { + "epoch": 4.803514376996805, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 6014 + }, + { + "epoch": 4.804313099041534, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 6015 + }, + { + "epoch": 4.805111821086262, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 6016 + }, + { + "epoch": 4.805910543130991, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 6017 + }, + { + "epoch": 4.806709265175719, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 6018 + }, + { + "epoch": 4.807507987220447, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 6019 + }, + { + "epoch": 4.8083067092651754, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 6020 + }, + { + "epoch": 4.809105431309904, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 6021 + }, + { + "epoch": 4.8099041533546325, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 6022 + }, + { + "epoch": 4.810702875399361, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 6023 + }, + { + "epoch": 4.81150159744409, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 6024 + }, + { + "epoch": 4.812300319488818, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 6025 + }, + { + "epoch": 4.813099041533547, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 6026 + }, + { + "epoch": 4.813897763578275, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 6027 + }, + { + "epoch": 4.814696485623003, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 6028 + }, + { + "epoch": 4.815495207667731, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 6029 + }, + { + "epoch": 4.81629392971246, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 6030 + }, + { + "epoch": 4.817092651757188, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 6031 + }, + { + "epoch": 4.817891373801917, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 6032 + }, + { + "epoch": 4.818690095846645, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 6033 + }, + { + "epoch": 4.819488817891374, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 6034 + }, + { + "epoch": 4.8202875399361025, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 6035 + }, + { + "epoch": 4.821086261980831, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 6036 + }, + { + "epoch": 4.821884984025559, + "grad_norm": 0.318359375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 6037 + }, + { + "epoch": 4.822683706070287, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 6038 + }, + { + "epoch": 4.823482428115016, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 6039 + }, + { + "epoch": 4.824281150159744, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 6040 + }, + { + "epoch": 4.825079872204473, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6041 + }, + { + "epoch": 4.825878594249201, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 6042 + }, + { + "epoch": 4.82667731629393, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 6043 + }, + { + "epoch": 4.827476038338658, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 6044 + }, + { + "epoch": 4.828274760383387, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 6045 + }, + { + "epoch": 4.8290734824281145, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 6046 + }, + { + "epoch": 4.829872204472844, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 6047 + }, + { + "epoch": 4.830670926517572, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 6048 + }, + { + "epoch": 4.8314696485623, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 6049 + }, + { + "epoch": 4.832268370607029, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 6050 + }, + { + "epoch": 4.833067092651757, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 6051 + }, + { + "epoch": 4.833865814696486, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 6052 + }, + { + "epoch": 4.834664536741214, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 6053 + }, + { + "epoch": 4.835463258785943, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 6054 + }, + { + "epoch": 4.836261980830671, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 6055 + }, + { + "epoch": 4.8370607028754, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 6056 + }, + { + "epoch": 4.837859424920127, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 6057 + }, + { + "epoch": 4.838658146964856, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 6058 + }, + { + "epoch": 4.8394568690095845, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 6059 + }, + { + "epoch": 4.840255591054313, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 6060 + }, + { + "epoch": 4.8410543130990416, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 6061 + }, + { + "epoch": 4.84185303514377, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 6062 + }, + { + "epoch": 4.842651757188499, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 6063 + }, + { + "epoch": 4.843450479233227, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 6064 + }, + { + "epoch": 4.844249201277956, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 6065 + }, + { + "epoch": 4.845047923322683, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 6066 + }, + { + "epoch": 4.845846645367412, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 6067 + }, + { + "epoch": 4.84664536741214, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 6068 + }, + { + "epoch": 4.847444089456869, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 6069 + }, + { + "epoch": 4.848242811501597, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 6070 + }, + { + "epoch": 4.849041533546326, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 6071 + }, + { + "epoch": 4.8498402555910545, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 6072 + }, + { + "epoch": 4.850638977635783, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 6073 + }, + { + "epoch": 4.8514376996805115, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 6074 + }, + { + "epoch": 4.852236421725239, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 6075 + }, + { + "epoch": 4.853035143769968, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1245, + "step": 6076 + }, + { + "epoch": 4.853833865814696, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 6077 + }, + { + "epoch": 4.854632587859425, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 6078 + }, + { + "epoch": 4.855431309904153, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 6079 + }, + { + "epoch": 4.856230031948882, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 6080 + }, + { + "epoch": 4.85702875399361, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 6081 + }, + { + "epoch": 4.857827476038339, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 6082 + }, + { + "epoch": 4.858626198083067, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 6083 + }, + { + "epoch": 4.859424920127795, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 6084 + }, + { + "epoch": 4.8602236421725244, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 6085 + }, + { + "epoch": 4.861022364217252, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 6086 + }, + { + "epoch": 4.861821086261981, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 6087 + }, + { + "epoch": 4.862619808306709, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 6088 + }, + { + "epoch": 4.863418530351438, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 6089 + }, + { + "epoch": 4.864217252396166, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 6090 + }, + { + "epoch": 4.865015974440895, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 6091 + }, + { + "epoch": 4.865814696485623, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 6092 + }, + { + "epoch": 4.866613418530352, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 6093 + }, + { + "epoch": 4.86741214057508, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 6094 + }, + { + "epoch": 4.868210862619808, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 6095 + }, + { + "epoch": 4.8690095846645365, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 6096 + }, + { + "epoch": 4.869808306709265, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 6097 + }, + { + "epoch": 4.8706070287539935, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 6098 + }, + { + "epoch": 4.871405750798722, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 6099 + }, + { + "epoch": 4.872204472843451, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1236, + "step": 6100 + }, + { + "epoch": 4.873003194888179, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 6101 + }, + { + "epoch": 4.873801916932908, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 6102 + }, + { + "epoch": 4.874600638977636, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 6103 + }, + { + "epoch": 4.875399361022364, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 6104 + }, + { + "epoch": 4.876198083067092, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 6105 + }, + { + "epoch": 4.876996805111821, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 6106 + }, + { + "epoch": 4.877795527156549, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 6107 + }, + { + "epoch": 4.878594249201278, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 6108 + }, + { + "epoch": 4.8793929712460065, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 6109 + }, + { + "epoch": 4.880191693290735, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 6110 + }, + { + "epoch": 4.8809904153354635, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1253, + "step": 6111 + }, + { + "epoch": 4.881789137380192, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 6112 + }, + { + "epoch": 4.88258785942492, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 6113 + }, + { + "epoch": 4.883386581469648, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 6114 + }, + { + "epoch": 4.884185303514377, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 6115 + }, + { + "epoch": 4.884984025559105, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 6116 + }, + { + "epoch": 4.885782747603834, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 6117 + }, + { + "epoch": 4.886581469648562, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 6118 + }, + { + "epoch": 4.887380191693291, + "grad_norm": 0.09814453125, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 6119 + }, + { + "epoch": 4.888178913738019, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 6120 + }, + { + "epoch": 4.888977635782748, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 6121 + }, + { + "epoch": 4.8897763578274756, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 6122 + }, + { + "epoch": 4.890575079872205, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 6123 + }, + { + "epoch": 4.891373801916933, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 6124 + }, + { + "epoch": 4.892172523961661, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 6125 + }, + { + "epoch": 4.89297124600639, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 6126 + }, + { + "epoch": 4.893769968051118, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 6127 + }, + { + "epoch": 4.894568690095847, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 6128 + }, + { + "epoch": 4.895367412140575, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 6129 + }, + { + "epoch": 4.896166134185304, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 6130 + }, + { + "epoch": 4.896964856230032, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 6131 + }, + { + "epoch": 4.897763578274761, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6132 + }, + { + "epoch": 4.8985623003194885, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 6133 + }, + { + "epoch": 4.899361022364217, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 6134 + }, + { + "epoch": 4.9001597444089455, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 6135 + }, + { + "epoch": 4.900958466453674, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6136 + }, + { + "epoch": 4.901757188498403, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 6137 + }, + { + "epoch": 4.902555910543131, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 6138 + }, + { + "epoch": 4.90335463258786, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 6139 + }, + { + "epoch": 4.904153354632588, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 6140 + }, + { + "epoch": 4.904952076677317, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 6141 + }, + { + "epoch": 4.905750798722044, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 6142 + }, + { + "epoch": 4.906549520766773, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 6143 + }, + { + "epoch": 4.907348242811501, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 6144 + }, + { + "epoch": 4.90814696485623, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 6145 + }, + { + "epoch": 4.9089456869009584, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 6146 + }, + { + "epoch": 4.909744408945687, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 6147 + }, + { + "epoch": 4.9105431309904155, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 6148 + }, + { + "epoch": 4.911341853035144, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 6149 + }, + { + "epoch": 4.912140575079873, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 6150 + }, + { + "epoch": 4.9129392971246, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 6151 + }, + { + "epoch": 4.913738019169329, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 6152 + }, + { + "epoch": 4.914536741214057, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 6153 + }, + { + "epoch": 4.915335463258786, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 6154 + }, + { + "epoch": 4.916134185303514, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 6155 + }, + { + "epoch": 4.916932907348243, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 6156 + }, + { + "epoch": 4.917731629392971, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 6157 + }, + { + "epoch": 4.9185303514377, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 6158 + }, + { + "epoch": 4.919329073482428, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 6159 + }, + { + "epoch": 4.920127795527156, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 6160 + }, + { + "epoch": 4.9209265175718855, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 6161 + }, + { + "epoch": 4.921725239616613, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 6162 + }, + { + "epoch": 4.922523961661342, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 6163 + }, + { + "epoch": 4.92332268370607, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1307, + "step": 6164 + }, + { + "epoch": 4.924121405750799, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 6165 + }, + { + "epoch": 4.924920127795527, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1222, + "step": 6166 + }, + { + "epoch": 4.925718849840256, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 6167 + }, + { + "epoch": 4.926517571884984, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 6168 + }, + { + "epoch": 4.927316293929713, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 6169 + }, + { + "epoch": 4.928115015974441, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 6170 + }, + { + "epoch": 4.928913738019169, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 6171 + }, + { + "epoch": 4.9297124600638975, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 6172 + }, + { + "epoch": 4.930511182108626, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.122, + "step": 6173 + }, + { + "epoch": 4.931309904153355, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 6174 + }, + { + "epoch": 4.932108626198083, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 6175 + }, + { + "epoch": 4.932907348242812, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 6176 + }, + { + "epoch": 4.93370607028754, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 6177 + }, + { + "epoch": 4.934504792332269, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 6178 + }, + { + "epoch": 4.935303514376997, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 6179 + }, + { + "epoch": 4.936102236421725, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 6180 + }, + { + "epoch": 4.936900958466453, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 6181 + }, + { + "epoch": 4.937699680511182, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 6182 + }, + { + "epoch": 4.93849840255591, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 6183 + }, + { + "epoch": 4.939297124600639, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 6184 + }, + { + "epoch": 4.9400958466453675, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 6185 + }, + { + "epoch": 4.940894568690096, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 6186 + }, + { + "epoch": 4.9416932907348246, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 6187 + }, + { + "epoch": 4.942492012779553, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 6188 + }, + { + "epoch": 4.943290734824281, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 6189 + }, + { + "epoch": 4.944089456869009, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 6190 + }, + { + "epoch": 4.944888178913738, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6191 + }, + { + "epoch": 4.945686900958466, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 6192 + }, + { + "epoch": 4.946485623003195, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 6193 + }, + { + "epoch": 4.947284345047923, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 6194 + }, + { + "epoch": 4.948083067092652, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 6195 + }, + { + "epoch": 4.94888178913738, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 6196 + }, + { + "epoch": 4.949680511182109, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 6197 + }, + { + "epoch": 4.950479233226837, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 6198 + }, + { + "epoch": 4.951277955271565, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 6199 + }, + { + "epoch": 4.952076677316294, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 6200 + }, + { + "epoch": 4.952875399361022, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 6201 + }, + { + "epoch": 4.953674121405751, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 6202 + }, + { + "epoch": 4.954472843450479, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 6203 + }, + { + "epoch": 4.955271565495208, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 6204 + }, + { + "epoch": 4.956070287539936, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 6205 + }, + { + "epoch": 4.956869009584665, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 6206 + }, + { + "epoch": 4.957667731629393, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 6207 + }, + { + "epoch": 4.958466453674122, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 6208 + }, + { + "epoch": 4.9592651757188495, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 6209 + }, + { + "epoch": 4.960063897763578, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 6210 + }, + { + "epoch": 4.960862619808307, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 6211 + }, + { + "epoch": 4.961661341853035, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 6212 + }, + { + "epoch": 4.962460063897764, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 6213 + }, + { + "epoch": 4.963258785942492, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 6214 + }, + { + "epoch": 4.964057507987221, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 6215 + }, + { + "epoch": 4.964856230031949, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 6216 + }, + { + "epoch": 4.965654952076678, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1235, + "step": 6217 + }, + { + "epoch": 4.966453674121405, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 6218 + }, + { + "epoch": 4.967252396166134, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 6219 + }, + { + "epoch": 4.968051118210862, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 6220 + }, + { + "epoch": 4.968849840255591, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1232, + "step": 6221 + }, + { + "epoch": 4.9696485623003195, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 6222 + }, + { + "epoch": 4.970447284345048, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 6223 + }, + { + "epoch": 4.9712460063897765, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 6224 + }, + { + "epoch": 4.972044728434505, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 6225 + }, + { + "epoch": 4.972843450479234, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 6226 + }, + { + "epoch": 4.973642172523961, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 6227 + }, + { + "epoch": 4.97444089456869, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 6228 + }, + { + "epoch": 4.975239616613418, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 6229 + }, + { + "epoch": 4.976038338658147, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 6230 + }, + { + "epoch": 4.976837060702875, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 6231 + }, + { + "epoch": 4.977635782747604, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 6232 + }, + { + "epoch": 4.978434504792332, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 6233 + }, + { + "epoch": 4.979233226837061, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 6234 + }, + { + "epoch": 4.9800319488817895, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 6235 + }, + { + "epoch": 4.980830670926517, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 6236 + }, + { + "epoch": 4.981629392971246, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 6237 + }, + { + "epoch": 4.982428115015974, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 6238 + }, + { + "epoch": 4.983226837060703, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 6239 + }, + { + "epoch": 4.984025559105431, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 6240 + }, + { + "epoch": 4.98482428115016, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 6241 + }, + { + "epoch": 4.985623003194888, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 6242 + }, + { + "epoch": 4.986421725239617, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 6243 + }, + { + "epoch": 4.987220447284345, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 6244 + }, + { + "epoch": 4.988019169329074, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 6245 + }, + { + "epoch": 4.988817891373802, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 6246 + }, + { + "epoch": 4.98961661341853, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 6247 + }, + { + "epoch": 4.9904153354632586, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 6248 + }, + { + "epoch": 4.991214057507987, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 6249 + }, + { + "epoch": 4.992012779552716, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1261, + "step": 6250 + }, + { + "epoch": 4.992811501597444, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 6251 + }, + { + "epoch": 4.993610223642173, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 6252 + }, + { + "epoch": 4.994408945686901, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 6253 + }, + { + "epoch": 4.99520766773163, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 6254 + }, + { + "epoch": 4.996006389776358, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 6255 + }, + { + "epoch": 4.996805111821086, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 6256 + }, + { + "epoch": 4.997603833865814, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 6257 + }, + { + "epoch": 4.998402555910543, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 6258 + }, + { + "epoch": 4.9992012779552715, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 6259 + }, + { + "epoch": 5.0, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6260 + }, + { + "epoch": 5.0007987220447285, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 6261 + }, + { + "epoch": 5.001597444089457, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 6262 + }, + { + "epoch": 5.002396166134186, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 6263 + }, + { + "epoch": 5.003194888178914, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 6264 + }, + { + "epoch": 5.003993610223642, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 6265 + }, + { + "epoch": 5.00479233226837, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 6266 + }, + { + "epoch": 5.005591054313099, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 6267 + }, + { + "epoch": 5.006389776357827, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 6268 + }, + { + "epoch": 5.007188498402556, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 6269 + }, + { + "epoch": 5.007987220447284, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 6270 + }, + { + "epoch": 5.008785942492013, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 6271 + }, + { + "epoch": 5.0095846645367414, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 6272 + }, + { + "epoch": 5.01038338658147, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 6273 + }, + { + "epoch": 5.0111821086261985, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 6274 + }, + { + "epoch": 5.011980830670926, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 6275 + }, + { + "epoch": 5.012779552715655, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 6276 + }, + { + "epoch": 5.013578274760383, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 6277 + }, + { + "epoch": 5.014376996805112, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 6278 + }, + { + "epoch": 5.01517571884984, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 6279 + }, + { + "epoch": 5.015974440894569, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 6280 + }, + { + "epoch": 5.016773162939297, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 6281 + }, + { + "epoch": 5.017571884984026, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 6282 + }, + { + "epoch": 5.018370607028754, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 6283 + }, + { + "epoch": 5.019169329073482, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 6284 + }, + { + "epoch": 5.0199680511182105, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 6285 + }, + { + "epoch": 5.020766773162939, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 6286 + }, + { + "epoch": 5.021565495207668, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 6287 + }, + { + "epoch": 5.022364217252396, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 6288 + }, + { + "epoch": 5.023162939297125, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 6289 + }, + { + "epoch": 5.023961661341853, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 6290 + }, + { + "epoch": 5.024760383386582, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 6291 + }, + { + "epoch": 5.02555910543131, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 6292 + }, + { + "epoch": 5.026357827476039, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 6293 + }, + { + "epoch": 5.027156549520766, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 6294 + }, + { + "epoch": 5.027955271565495, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 6295 + }, + { + "epoch": 5.0287539936102235, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 6296 + }, + { + "epoch": 5.029552715654952, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 6297 + }, + { + "epoch": 5.0303514376996805, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 6298 + }, + { + "epoch": 5.031150159744409, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 6299 + }, + { + "epoch": 5.031948881789138, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 6300 + }, + { + "epoch": 5.032747603833866, + "grad_norm": 0.408203125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 6301 + }, + { + "epoch": 5.033546325878595, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 6302 + }, + { + "epoch": 5.034345047923322, + "grad_norm": 0.421875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 6303 + }, + { + "epoch": 5.035143769968051, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 6304 + }, + { + "epoch": 5.035942492012779, + "grad_norm": 0.443359375, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 6305 + }, + { + "epoch": 5.036741214057508, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 6306 + }, + { + "epoch": 5.037539936102236, + "grad_norm": 0.51171875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6307 + }, + { + "epoch": 5.038338658146965, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 6308 + }, + { + "epoch": 5.039137380191693, + "grad_norm": 0.55859375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 6309 + }, + { + "epoch": 5.039936102236422, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 6310 + }, + { + "epoch": 5.0407348242811505, + "grad_norm": 0.54296875, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 6311 + }, + { + "epoch": 5.041533546325879, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 6312 + }, + { + "epoch": 5.042332268370607, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.1218, + "step": 6313 + }, + { + "epoch": 5.043130990415335, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 6314 + }, + { + "epoch": 5.043929712460064, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 6315 + }, + { + "epoch": 5.044728434504792, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 6316 + }, + { + "epoch": 5.045527156549521, + "grad_norm": 0.421875, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 6317 + }, + { + "epoch": 5.046325878594249, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1289, + "step": 6318 + }, + { + "epoch": 5.047124600638978, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 6319 + }, + { + "epoch": 5.047923322683706, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 6320 + }, + { + "epoch": 5.048722044728435, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 6321 + }, + { + "epoch": 5.0495207667731625, + "grad_norm": 0.345703125, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 6322 + }, + { + "epoch": 5.050319488817891, + "grad_norm": 0.37109375, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 6323 + }, + { + "epoch": 5.05111821086262, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 6324 + }, + { + "epoch": 5.051916932907348, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 6325 + }, + { + "epoch": 5.052715654952077, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 6326 + }, + { + "epoch": 5.053514376996805, + "grad_norm": 0.494140625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 6327 + }, + { + "epoch": 5.054313099041534, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 6328 + }, + { + "epoch": 5.055111821086262, + "grad_norm": 0.94140625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 6329 + }, + { + "epoch": 5.055910543130991, + "grad_norm": 0.45703125, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 6330 + }, + { + "epoch": 5.056709265175719, + "grad_norm": 0.51171875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 6331 + }, + { + "epoch": 5.057507987220447, + "grad_norm": 0.84375, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 6332 + }, + { + "epoch": 5.0583067092651754, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 6333 + }, + { + "epoch": 5.059105431309904, + "grad_norm": 0.625, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 6334 + }, + { + "epoch": 5.0599041533546325, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 6335 + }, + { + "epoch": 5.060702875399361, + "grad_norm": 0.515625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 6336 + }, + { + "epoch": 5.06150159744409, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 6337 + }, + { + "epoch": 5.062300319488818, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 6338 + }, + { + "epoch": 5.063099041533547, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 6339 + }, + { + "epoch": 5.063897763578275, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 6340 + }, + { + "epoch": 5.064696485623003, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 6341 + }, + { + "epoch": 5.065495207667731, + "grad_norm": 0.353515625, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 6342 + }, + { + "epoch": 5.06629392971246, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 6343 + }, + { + "epoch": 5.067092651757188, + "grad_norm": 0.39453125, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 6344 + }, + { + "epoch": 5.067891373801917, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 6345 + }, + { + "epoch": 5.068690095846645, + "grad_norm": 0.50390625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 6346 + }, + { + "epoch": 5.069488817891374, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 6347 + }, + { + "epoch": 5.0702875399361025, + "grad_norm": 0.515625, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 6348 + }, + { + "epoch": 5.071086261980831, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 6349 + }, + { + "epoch": 5.0718849840255595, + "grad_norm": 0.5625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 6350 + }, + { + "epoch": 5.072683706070287, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 6351 + }, + { + "epoch": 5.073482428115016, + "grad_norm": 0.64453125, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 6352 + }, + { + "epoch": 5.074281150159744, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1193, + "step": 6353 + }, + { + "epoch": 5.075079872204473, + "grad_norm": 0.7109375, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 6354 + }, + { + "epoch": 5.075878594249201, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 6355 + }, + { + "epoch": 5.07667731629393, + "grad_norm": 0.71484375, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 6356 + }, + { + "epoch": 5.077476038338658, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 6357 + }, + { + "epoch": 5.078274760383387, + "grad_norm": 0.57421875, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 6358 + }, + { + "epoch": 5.079073482428115, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 6359 + }, + { + "epoch": 5.079872204472843, + "grad_norm": 0.4140625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 6360 + }, + { + "epoch": 5.080670926517572, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 6361 + }, + { + "epoch": 5.0814696485623, + "grad_norm": 0.375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 6362 + }, + { + "epoch": 5.082268370607029, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6363 + }, + { + "epoch": 5.083067092651757, + "grad_norm": 0.369140625, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 6364 + }, + { + "epoch": 5.083865814696486, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 6365 + }, + { + "epoch": 5.084664536741214, + "grad_norm": 0.39453125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 6366 + }, + { + "epoch": 5.085463258785943, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 6367 + }, + { + "epoch": 5.086261980830671, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 6368 + }, + { + "epoch": 5.0870607028754, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 6369 + }, + { + "epoch": 5.087859424920127, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 6370 + }, + { + "epoch": 5.088658146964856, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 6371 + }, + { + "epoch": 5.0894568690095845, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 6372 + }, + { + "epoch": 5.090255591054313, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 6373 + }, + { + "epoch": 5.0910543130990416, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 6374 + }, + { + "epoch": 5.09185303514377, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 6375 + }, + { + "epoch": 5.092651757188499, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 6376 + }, + { + "epoch": 5.093450479233227, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.124, + "step": 6377 + }, + { + "epoch": 5.094249201277956, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 6378 + }, + { + "epoch": 5.095047923322683, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 6379 + }, + { + "epoch": 5.095846645367412, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 6380 + }, + { + "epoch": 5.09664536741214, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 6381 + }, + { + "epoch": 5.097444089456869, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 6382 + }, + { + "epoch": 5.098242811501597, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 6383 + }, + { + "epoch": 5.099041533546326, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 6384 + }, + { + "epoch": 5.0998402555910545, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 6385 + }, + { + "epoch": 5.100638977635783, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 6386 + }, + { + "epoch": 5.1014376996805115, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 6387 + }, + { + "epoch": 5.102236421725239, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 6388 + }, + { + "epoch": 5.103035143769968, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 6389 + }, + { + "epoch": 5.103833865814696, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 6390 + }, + { + "epoch": 5.104632587859425, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 6391 + }, + { + "epoch": 5.105431309904153, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 6392 + }, + { + "epoch": 5.106230031948882, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 6393 + }, + { + "epoch": 5.10702875399361, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 6394 + }, + { + "epoch": 5.107827476038339, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 6395 + }, + { + "epoch": 5.108626198083067, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 6396 + }, + { + "epoch": 5.109424920127796, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 6397 + }, + { + "epoch": 5.110223642172524, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 6398 + }, + { + "epoch": 5.111022364217252, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 6399 + }, + { + "epoch": 5.111821086261981, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 6400 + }, + { + "epoch": 5.112619808306709, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 6401 + }, + { + "epoch": 5.113418530351438, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 6402 + }, + { + "epoch": 5.114217252396166, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 6403 + }, + { + "epoch": 5.115015974440895, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 6404 + }, + { + "epoch": 5.115814696485623, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 6405 + }, + { + "epoch": 5.116613418530352, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 6406 + }, + { + "epoch": 5.11741214057508, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 6407 + }, + { + "epoch": 5.118210862619808, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 6408 + }, + { + "epoch": 5.1190095846645365, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 6409 + }, + { + "epoch": 5.119808306709265, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 6410 + }, + { + "epoch": 5.1206070287539935, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 6411 + }, + { + "epoch": 5.121405750798722, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 6412 + }, + { + "epoch": 5.122204472843451, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 6413 + }, + { + "epoch": 5.123003194888179, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 6414 + }, + { + "epoch": 5.123801916932908, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 6415 + }, + { + "epoch": 5.124600638977636, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 6416 + }, + { + "epoch": 5.125399361022364, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 6417 + }, + { + "epoch": 5.126198083067092, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 6418 + }, + { + "epoch": 5.126996805111821, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 6419 + }, + { + "epoch": 5.127795527156549, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 6420 + }, + { + "epoch": 5.128594249201278, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 6421 + }, + { + "epoch": 5.1293929712460065, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 6422 + }, + { + "epoch": 5.130191693290735, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 6423 + }, + { + "epoch": 5.1309904153354635, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 6424 + }, + { + "epoch": 5.131789137380192, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 6425 + }, + { + "epoch": 5.13258785942492, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 6426 + }, + { + "epoch": 5.133386581469648, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 6427 + }, + { + "epoch": 5.134185303514377, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 6428 + }, + { + "epoch": 5.134984025559105, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 6429 + }, + { + "epoch": 5.135782747603834, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 6430 + }, + { + "epoch": 5.136581469648562, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 6431 + }, + { + "epoch": 5.137380191693291, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 6432 + }, + { + "epoch": 5.138178913738019, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 6433 + }, + { + "epoch": 5.138977635782748, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 6434 + }, + { + "epoch": 5.139776357827476, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 6435 + }, + { + "epoch": 5.140575079872204, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 6436 + }, + { + "epoch": 5.141373801916933, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 6437 + }, + { + "epoch": 5.142172523961661, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 6438 + }, + { + "epoch": 5.14297124600639, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 6439 + }, + { + "epoch": 5.143769968051118, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 6440 + }, + { + "epoch": 5.144568690095847, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 6441 + }, + { + "epoch": 5.145367412140575, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 6442 + }, + { + "epoch": 5.146166134185304, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 6443 + }, + { + "epoch": 5.146964856230032, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 6444 + }, + { + "epoch": 5.147763578274761, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 6445 + }, + { + "epoch": 5.1485623003194885, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 6446 + }, + { + "epoch": 5.149361022364217, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 6447 + }, + { + "epoch": 5.1501597444089455, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1246, + "step": 6448 + }, + { + "epoch": 5.150958466453674, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 6449 + }, + { + "epoch": 5.151757188498403, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 6450 + }, + { + "epoch": 5.152555910543131, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 6451 + }, + { + "epoch": 5.15335463258786, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 6452 + }, + { + "epoch": 5.154153354632588, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 6453 + }, + { + "epoch": 5.154952076677317, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 6454 + }, + { + "epoch": 5.155750798722044, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 6455 + }, + { + "epoch": 5.156549520766773, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 6456 + }, + { + "epoch": 5.157348242811501, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 6457 + }, + { + "epoch": 5.15814696485623, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 6458 + }, + { + "epoch": 5.1589456869009584, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 6459 + }, + { + "epoch": 5.159744408945687, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 6460 + }, + { + "epoch": 5.1605431309904155, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 6461 + }, + { + "epoch": 5.161341853035144, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 6462 + }, + { + "epoch": 5.162140575079873, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 6463 + }, + { + "epoch": 5.1629392971246, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 6464 + }, + { + "epoch": 5.163738019169329, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 6465 + }, + { + "epoch": 5.164536741214057, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 6466 + }, + { + "epoch": 5.165335463258786, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6467 + }, + { + "epoch": 5.166134185303514, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 6468 + }, + { + "epoch": 5.166932907348243, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 6469 + }, + { + "epoch": 5.167731629392971, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 6470 + }, + { + "epoch": 5.1685303514377, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 6471 + }, + { + "epoch": 5.169329073482428, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 6472 + }, + { + "epoch": 5.170127795527157, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 6473 + }, + { + "epoch": 5.170926517571885, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 6474 + }, + { + "epoch": 5.171725239616613, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 6475 + }, + { + "epoch": 5.172523961661342, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 6476 + }, + { + "epoch": 5.17332268370607, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 6477 + }, + { + "epoch": 5.174121405750799, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 6478 + }, + { + "epoch": 5.174920127795527, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6479 + }, + { + "epoch": 5.175718849840256, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 6480 + }, + { + "epoch": 5.176517571884984, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1225, + "step": 6481 + }, + { + "epoch": 5.177316293929713, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 6482 + }, + { + "epoch": 5.178115015974441, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 6483 + }, + { + "epoch": 5.178913738019169, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1242, + "step": 6484 + }, + { + "epoch": 5.1797124600638975, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 6485 + }, + { + "epoch": 5.180511182108626, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 6486 + }, + { + "epoch": 5.181309904153355, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 6487 + }, + { + "epoch": 5.182108626198083, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 6488 + }, + { + "epoch": 5.182907348242812, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 6489 + }, + { + "epoch": 5.18370607028754, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 6490 + }, + { + "epoch": 5.184504792332269, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 6491 + }, + { + "epoch": 5.185303514376997, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 6492 + }, + { + "epoch": 5.186102236421725, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 6493 + }, + { + "epoch": 5.186900958466453, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 6494 + }, + { + "epoch": 5.187699680511182, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 6495 + }, + { + "epoch": 5.18849840255591, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 6496 + }, + { + "epoch": 5.189297124600639, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 6497 + }, + { + "epoch": 5.1900958466453675, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 6498 + }, + { + "epoch": 5.190894568690096, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 6499 + }, + { + "epoch": 5.1916932907348246, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 6500 + }, + { + "epoch": 5.192492012779553, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 6501 + }, + { + "epoch": 5.193290734824281, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 6502 + }, + { + "epoch": 5.194089456869009, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 6503 + }, + { + "epoch": 5.194888178913738, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 6504 + }, + { + "epoch": 5.195686900958466, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 6505 + }, + { + "epoch": 5.196485623003195, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 6506 + }, + { + "epoch": 5.197284345047923, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 6507 + }, + { + "epoch": 5.198083067092652, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 6508 + }, + { + "epoch": 5.19888178913738, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 6509 + }, + { + "epoch": 5.199680511182109, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 6510 + }, + { + "epoch": 5.2004792332268375, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 6511 + }, + { + "epoch": 5.201277955271565, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 6512 + }, + { + "epoch": 5.202076677316294, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 6513 + }, + { + "epoch": 5.202875399361022, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 6514 + }, + { + "epoch": 5.203674121405751, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 6515 + }, + { + "epoch": 5.204472843450479, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 6516 + }, + { + "epoch": 5.205271565495208, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 6517 + }, + { + "epoch": 5.206070287539936, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 6518 + }, + { + "epoch": 5.206869009584665, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 6519 + }, + { + "epoch": 5.207667731629393, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 6520 + }, + { + "epoch": 5.208466453674121, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 6521 + }, + { + "epoch": 5.2092651757188495, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 6522 + }, + { + "epoch": 5.210063897763578, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 6523 + }, + { + "epoch": 5.210862619808307, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 6524 + }, + { + "epoch": 5.211661341853035, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 6525 + }, + { + "epoch": 5.212460063897764, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 6526 + }, + { + "epoch": 5.213258785942492, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 6527 + }, + { + "epoch": 5.214057507987221, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 6528 + }, + { + "epoch": 5.214856230031949, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1252, + "step": 6529 + }, + { + "epoch": 5.215654952076678, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 6530 + }, + { + "epoch": 5.216453674121405, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 6531 + }, + { + "epoch": 5.217252396166134, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 6532 + }, + { + "epoch": 5.218051118210862, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 6533 + }, + { + "epoch": 5.218849840255591, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 6534 + }, + { + "epoch": 5.2196485623003195, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 6535 + }, + { + "epoch": 5.220447284345048, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 6536 + }, + { + "epoch": 5.2212460063897765, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 6537 + }, + { + "epoch": 5.222044728434505, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 6538 + }, + { + "epoch": 5.222843450479234, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 6539 + }, + { + "epoch": 5.223642172523961, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 6540 + }, + { + "epoch": 5.22444089456869, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 6541 + }, + { + "epoch": 5.225239616613418, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 6542 + }, + { + "epoch": 5.226038338658147, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 6543 + }, + { + "epoch": 5.226837060702875, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 6544 + }, + { + "epoch": 5.227635782747604, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 6545 + }, + { + "epoch": 5.228434504792332, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 6546 + }, + { + "epoch": 5.229233226837061, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 6547 + }, + { + "epoch": 5.2300319488817895, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6548 + }, + { + "epoch": 5.230830670926518, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 6549 + }, + { + "epoch": 5.231629392971246, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 6550 + }, + { + "epoch": 5.232428115015974, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 6551 + }, + { + "epoch": 5.233226837060703, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 6552 + }, + { + "epoch": 5.234025559105431, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 6553 + }, + { + "epoch": 5.23482428115016, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 6554 + }, + { + "epoch": 5.235623003194888, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 6555 + }, + { + "epoch": 5.236421725239617, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 6556 + }, + { + "epoch": 5.237220447284345, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 6557 + }, + { + "epoch": 5.238019169329074, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 6558 + }, + { + "epoch": 5.2388178913738015, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 6559 + }, + { + "epoch": 5.23961661341853, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 6560 + }, + { + "epoch": 5.2404153354632586, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 6561 + }, + { + "epoch": 5.241214057507987, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 6562 + }, + { + "epoch": 5.242012779552716, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 6563 + }, + { + "epoch": 5.242811501597444, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 6564 + }, + { + "epoch": 5.243610223642173, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 6565 + }, + { + "epoch": 5.244408945686901, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 6566 + }, + { + "epoch": 5.24520766773163, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 6567 + }, + { + "epoch": 5.246006389776358, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 6568 + }, + { + "epoch": 5.246805111821086, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 6569 + }, + { + "epoch": 5.247603833865814, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 6570 + }, + { + "epoch": 5.248402555910543, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 6571 + }, + { + "epoch": 5.2492012779552715, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 6572 + }, + { + "epoch": 5.25, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 6573 + }, + { + "epoch": 5.2507987220447285, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 6574 + }, + { + "epoch": 5.251597444089457, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6575 + }, + { + "epoch": 5.252396166134186, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 6576 + }, + { + "epoch": 5.253194888178914, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 6577 + }, + { + "epoch": 5.253993610223642, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 6578 + }, + { + "epoch": 5.25479233226837, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 6579 + }, + { + "epoch": 5.255591054313099, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 6580 + }, + { + "epoch": 5.256389776357827, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 6581 + }, + { + "epoch": 5.257188498402556, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 6582 + }, + { + "epoch": 5.257987220447284, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 6583 + }, + { + "epoch": 5.258785942492013, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 6584 + }, + { + "epoch": 5.2595846645367414, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 6585 + }, + { + "epoch": 5.26038338658147, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 6586 + }, + { + "epoch": 5.261182108626198, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 6587 + }, + { + "epoch": 5.261980830670926, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 6588 + }, + { + "epoch": 5.262779552715655, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 6589 + }, + { + "epoch": 5.263578274760383, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 6590 + }, + { + "epoch": 5.264376996805112, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 6591 + }, + { + "epoch": 5.26517571884984, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 6592 + }, + { + "epoch": 5.265974440894569, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 6593 + }, + { + "epoch": 5.266773162939297, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 6594 + }, + { + "epoch": 5.267571884984026, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 6595 + }, + { + "epoch": 5.268370607028754, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 6596 + }, + { + "epoch": 5.269169329073483, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 6597 + }, + { + "epoch": 5.2699680511182105, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 6598 + }, + { + "epoch": 5.270766773162939, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 6599 + }, + { + "epoch": 5.271565495207668, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 6600 + }, + { + "epoch": 5.272364217252396, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 6601 + }, + { + "epoch": 5.273162939297125, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 6602 + }, + { + "epoch": 5.273961661341853, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 6603 + }, + { + "epoch": 5.274760383386582, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 6604 + }, + { + "epoch": 5.27555910543131, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.118, + "step": 6605 + }, + { + "epoch": 5.276357827476039, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 6606 + }, + { + "epoch": 5.277156549520766, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 6607 + }, + { + "epoch": 5.277955271565495, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 6608 + }, + { + "epoch": 5.2787539936102235, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 6609 + }, + { + "epoch": 5.279552715654952, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 6610 + }, + { + "epoch": 5.2803514376996805, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 6611 + }, + { + "epoch": 5.281150159744409, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 6612 + }, + { + "epoch": 5.281948881789138, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 6613 + }, + { + "epoch": 5.282747603833866, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 6614 + }, + { + "epoch": 5.283546325878595, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 6615 + }, + { + "epoch": 5.284345047923322, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 6616 + }, + { + "epoch": 5.285143769968051, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 6617 + }, + { + "epoch": 5.285942492012779, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 6618 + }, + { + "epoch": 5.286741214057508, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 6619 + }, + { + "epoch": 5.287539936102236, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 6620 + }, + { + "epoch": 5.288338658146965, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 6621 + }, + { + "epoch": 5.289137380191693, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 6622 + }, + { + "epoch": 5.289936102236422, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 6623 + }, + { + "epoch": 5.2907348242811505, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 6624 + }, + { + "epoch": 5.291533546325878, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 6625 + }, + { + "epoch": 5.292332268370607, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 6626 + }, + { + "epoch": 5.293130990415335, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 6627 + }, + { + "epoch": 5.293929712460064, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 6628 + }, + { + "epoch": 5.294728434504792, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 6629 + }, + { + "epoch": 5.295527156549521, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 6630 + }, + { + "epoch": 5.296325878594249, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 6631 + }, + { + "epoch": 5.297124600638978, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 6632 + }, + { + "epoch": 5.297923322683706, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 6633 + }, + { + "epoch": 5.298722044728435, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 6634 + }, + { + "epoch": 5.2995207667731625, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 6635 + }, + { + "epoch": 5.300319488817891, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 6636 + }, + { + "epoch": 5.30111821086262, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 6637 + }, + { + "epoch": 5.301916932907348, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 6638 + }, + { + "epoch": 5.302715654952077, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 6639 + }, + { + "epoch": 5.303514376996805, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6640 + }, + { + "epoch": 5.304313099041534, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 6641 + }, + { + "epoch": 5.305111821086262, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 6642 + }, + { + "epoch": 5.305910543130991, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 6643 + }, + { + "epoch": 5.306709265175719, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 6644 + }, + { + "epoch": 5.307507987220447, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 6645 + }, + { + "epoch": 5.3083067092651754, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 6646 + }, + { + "epoch": 5.309105431309904, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 6647 + }, + { + "epoch": 5.3099041533546325, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 6648 + }, + { + "epoch": 5.310702875399361, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 6649 + }, + { + "epoch": 5.31150159744409, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 6650 + }, + { + "epoch": 5.312300319488818, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 6651 + }, + { + "epoch": 5.313099041533547, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 6652 + }, + { + "epoch": 5.313897763578275, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 6653 + }, + { + "epoch": 5.314696485623003, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 6654 + }, + { + "epoch": 5.315495207667731, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 6655 + }, + { + "epoch": 5.31629392971246, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 6656 + }, + { + "epoch": 5.317092651757188, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 6657 + }, + { + "epoch": 5.317891373801917, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 6658 + }, + { + "epoch": 5.318690095846645, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 6659 + }, + { + "epoch": 5.319488817891374, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 6660 + }, + { + "epoch": 5.3202875399361025, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 6661 + }, + { + "epoch": 5.321086261980831, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 6662 + }, + { + "epoch": 5.321884984025559, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 6663 + }, + { + "epoch": 5.322683706070287, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 6664 + }, + { + "epoch": 5.323482428115016, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 6665 + }, + { + "epoch": 5.324281150159744, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 6666 + }, + { + "epoch": 5.325079872204473, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 6667 + }, + { + "epoch": 5.325878594249201, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 6668 + }, + { + "epoch": 5.32667731629393, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 6669 + }, + { + "epoch": 5.327476038338658, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 6670 + }, + { + "epoch": 5.328274760383387, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 6671 + }, + { + "epoch": 5.329073482428115, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 6672 + }, + { + "epoch": 5.329872204472843, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 6673 + }, + { + "epoch": 5.330670926517572, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 6674 + }, + { + "epoch": 5.3314696485623, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 6675 + }, + { + "epoch": 5.332268370607029, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 6676 + }, + { + "epoch": 5.333067092651757, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 6677 + }, + { + "epoch": 5.333865814696486, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 6678 + }, + { + "epoch": 5.334664536741214, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 6679 + }, + { + "epoch": 5.335463258785943, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 6680 + }, + { + "epoch": 5.336261980830671, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 6681 + }, + { + "epoch": 5.3370607028754, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 6682 + }, + { + "epoch": 5.337859424920127, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 6683 + }, + { + "epoch": 5.338658146964856, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 6684 + }, + { + "epoch": 5.3394568690095845, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 6685 + }, + { + "epoch": 5.340255591054313, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 6686 + }, + { + "epoch": 5.3410543130990416, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 6687 + }, + { + "epoch": 5.34185303514377, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 6688 + }, + { + "epoch": 5.342651757188499, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 6689 + }, + { + "epoch": 5.343450479233227, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 6690 + }, + { + "epoch": 5.344249201277956, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 6691 + }, + { + "epoch": 5.345047923322683, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 6692 + }, + { + "epoch": 5.345846645367412, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 6693 + }, + { + "epoch": 5.34664536741214, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 6694 + }, + { + "epoch": 5.347444089456869, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 6695 + }, + { + "epoch": 5.348242811501597, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6696 + }, + { + "epoch": 5.349041533546326, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 6697 + }, + { + "epoch": 5.3498402555910545, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 6698 + }, + { + "epoch": 5.350638977635783, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 6699 + }, + { + "epoch": 5.3514376996805115, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 6700 + }, + { + "epoch": 5.352236421725239, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 6701 + }, + { + "epoch": 5.353035143769968, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 6702 + }, + { + "epoch": 5.353833865814696, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 6703 + }, + { + "epoch": 5.354632587859425, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 6704 + }, + { + "epoch": 5.355431309904153, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 6705 + }, + { + "epoch": 5.356230031948882, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 6706 + }, + { + "epoch": 5.35702875399361, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 6707 + }, + { + "epoch": 5.357827476038339, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 6708 + }, + { + "epoch": 5.358626198083067, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 6709 + }, + { + "epoch": 5.359424920127796, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 6710 + }, + { + "epoch": 5.360223642172524, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 6711 + }, + { + "epoch": 5.361022364217252, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 6712 + }, + { + "epoch": 5.361821086261981, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 6713 + }, + { + "epoch": 5.362619808306709, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 6714 + }, + { + "epoch": 5.363418530351438, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 6715 + }, + { + "epoch": 5.364217252396166, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 6716 + }, + { + "epoch": 5.365015974440895, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 6717 + }, + { + "epoch": 5.365814696485623, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 6718 + }, + { + "epoch": 5.366613418530352, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 6719 + }, + { + "epoch": 5.36741214057508, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6720 + }, + { + "epoch": 5.368210862619808, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 6721 + }, + { + "epoch": 5.3690095846645365, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 6722 + }, + { + "epoch": 5.369808306709265, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 6723 + }, + { + "epoch": 5.3706070287539935, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 6724 + }, + { + "epoch": 5.371405750798722, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 6725 + }, + { + "epoch": 5.372204472843451, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 6726 + }, + { + "epoch": 5.373003194888179, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 6727 + }, + { + "epoch": 5.373801916932908, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 6728 + }, + { + "epoch": 5.374600638977636, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 6729 + }, + { + "epoch": 5.375399361022364, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 6730 + }, + { + "epoch": 5.376198083067092, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 6731 + }, + { + "epoch": 5.376996805111821, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 6732 + }, + { + "epoch": 5.377795527156549, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 6733 + }, + { + "epoch": 5.378594249201278, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 6734 + }, + { + "epoch": 5.3793929712460065, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 6735 + }, + { + "epoch": 5.380191693290735, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 6736 + }, + { + "epoch": 5.3809904153354635, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 6737 + }, + { + "epoch": 5.381789137380192, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 6738 + }, + { + "epoch": 5.38258785942492, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 6739 + }, + { + "epoch": 5.383386581469648, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 6740 + }, + { + "epoch": 5.384185303514377, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 6741 + }, + { + "epoch": 5.384984025559105, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 6742 + }, + { + "epoch": 5.385782747603834, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 6743 + }, + { + "epoch": 5.386581469648562, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 6744 + }, + { + "epoch": 5.387380191693291, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 6745 + }, + { + "epoch": 5.388178913738019, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 6746 + }, + { + "epoch": 5.388977635782748, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 6747 + }, + { + "epoch": 5.389776357827476, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 6748 + }, + { + "epoch": 5.390575079872204, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6749 + }, + { + "epoch": 5.391373801916933, + "grad_norm": 0.37109375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 6750 + }, + { + "epoch": 5.392172523961661, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 6751 + }, + { + "epoch": 5.39297124600639, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 6752 + }, + { + "epoch": 5.393769968051118, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6753 + }, + { + "epoch": 5.394568690095847, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 6754 + }, + { + "epoch": 5.395367412140575, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 6755 + }, + { + "epoch": 5.396166134185304, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 6756 + }, + { + "epoch": 5.396964856230032, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 6757 + }, + { + "epoch": 5.397763578274761, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 6758 + }, + { + "epoch": 5.3985623003194885, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 6759 + }, + { + "epoch": 5.399361022364217, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 6760 + }, + { + "epoch": 5.4001597444089455, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 6761 + }, + { + "epoch": 5.400958466453674, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 6762 + }, + { + "epoch": 5.401757188498403, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 6763 + }, + { + "epoch": 5.402555910543131, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 6764 + }, + { + "epoch": 5.40335463258786, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 6765 + }, + { + "epoch": 5.404153354632588, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 6766 + }, + { + "epoch": 5.404952076677317, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 6767 + }, + { + "epoch": 5.405750798722044, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 6768 + }, + { + "epoch": 5.406549520766773, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 6769 + }, + { + "epoch": 5.407348242811501, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 6770 + }, + { + "epoch": 5.40814696485623, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 6771 + }, + { + "epoch": 5.4089456869009584, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 6772 + }, + { + "epoch": 5.409744408945687, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 6773 + }, + { + "epoch": 5.4105431309904155, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 6774 + }, + { + "epoch": 5.411341853035144, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 6775 + }, + { + "epoch": 5.412140575079873, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 6776 + }, + { + "epoch": 5.4129392971246, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 6777 + }, + { + "epoch": 5.413738019169329, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 6778 + }, + { + "epoch": 5.414536741214057, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1214, + "step": 6779 + }, + { + "epoch": 5.415335463258786, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 6780 + }, + { + "epoch": 5.416134185303514, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 6781 + }, + { + "epoch": 5.416932907348243, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 6782 + }, + { + "epoch": 5.417731629392971, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 6783 + }, + { + "epoch": 5.4185303514377, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 6784 + }, + { + "epoch": 5.419329073482428, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 6785 + }, + { + "epoch": 5.420127795527157, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 6786 + }, + { + "epoch": 5.420926517571885, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 6787 + }, + { + "epoch": 5.421725239616613, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 6788 + }, + { + "epoch": 5.422523961661342, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 6789 + }, + { + "epoch": 5.42332268370607, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 6790 + }, + { + "epoch": 5.424121405750799, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 6791 + }, + { + "epoch": 5.424920127795527, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 6792 + }, + { + "epoch": 5.425718849840256, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 6793 + }, + { + "epoch": 5.426517571884984, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 6794 + }, + { + "epoch": 5.427316293929713, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 6795 + }, + { + "epoch": 5.428115015974441, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 6796 + }, + { + "epoch": 5.428913738019169, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 6797 + }, + { + "epoch": 5.4297124600638975, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 6798 + }, + { + "epoch": 5.430511182108626, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 6799 + }, + { + "epoch": 5.431309904153355, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 6800 + }, + { + "epoch": 5.432108626198083, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 6801 + }, + { + "epoch": 5.432907348242812, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 6802 + }, + { + "epoch": 5.43370607028754, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 6803 + }, + { + "epoch": 5.434504792332269, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 6804 + }, + { + "epoch": 5.435303514376997, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 6805 + }, + { + "epoch": 5.436102236421725, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 6806 + }, + { + "epoch": 5.436900958466453, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 6807 + }, + { + "epoch": 5.437699680511182, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 6808 + }, + { + "epoch": 5.43849840255591, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 6809 + }, + { + "epoch": 5.439297124600639, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 6810 + }, + { + "epoch": 5.4400958466453675, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 6811 + }, + { + "epoch": 5.440894568690096, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 6812 + }, + { + "epoch": 5.4416932907348246, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 6813 + }, + { + "epoch": 5.442492012779553, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 6814 + }, + { + "epoch": 5.443290734824281, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 6815 + }, + { + "epoch": 5.444089456869009, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 6816 + }, + { + "epoch": 5.444888178913738, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 6817 + }, + { + "epoch": 5.445686900958466, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 6818 + }, + { + "epoch": 5.446485623003195, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 6819 + }, + { + "epoch": 5.447284345047923, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 6820 + }, + { + "epoch": 5.448083067092652, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 6821 + }, + { + "epoch": 5.44888178913738, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6822 + }, + { + "epoch": 5.449680511182109, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 6823 + }, + { + "epoch": 5.4504792332268375, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 6824 + }, + { + "epoch": 5.451277955271565, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 6825 + }, + { + "epoch": 5.452076677316294, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 6826 + }, + { + "epoch": 5.452875399361022, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 6827 + }, + { + "epoch": 5.453674121405751, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 6828 + }, + { + "epoch": 5.454472843450479, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 6829 + }, + { + "epoch": 5.455271565495208, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 6830 + }, + { + "epoch": 5.456070287539936, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 6831 + }, + { + "epoch": 5.456869009584665, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 6832 + }, + { + "epoch": 5.457667731629393, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 6833 + }, + { + "epoch": 5.458466453674122, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 6834 + }, + { + "epoch": 5.4592651757188495, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 6835 + }, + { + "epoch": 5.460063897763578, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 6836 + }, + { + "epoch": 5.460862619808307, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 6837 + }, + { + "epoch": 5.461661341853035, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 6838 + }, + { + "epoch": 5.462460063897764, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 6839 + }, + { + "epoch": 5.463258785942492, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 6840 + }, + { + "epoch": 5.464057507987221, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 6841 + }, + { + "epoch": 5.464856230031949, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 6842 + }, + { + "epoch": 5.465654952076678, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 6843 + }, + { + "epoch": 5.466453674121405, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 6844 + }, + { + "epoch": 5.467252396166134, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 6845 + }, + { + "epoch": 5.468051118210862, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 6846 + }, + { + "epoch": 5.468849840255591, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 6847 + }, + { + "epoch": 5.4696485623003195, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 6848 + }, + { + "epoch": 5.470447284345048, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 6849 + }, + { + "epoch": 5.4712460063897765, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 6850 + }, + { + "epoch": 5.472044728434505, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 6851 + }, + { + "epoch": 5.472843450479234, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 6852 + }, + { + "epoch": 5.473642172523961, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 6853 + }, + { + "epoch": 5.47444089456869, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 6854 + }, + { + "epoch": 5.475239616613418, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 6855 + }, + { + "epoch": 5.476038338658147, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 6856 + }, + { + "epoch": 5.476837060702875, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 6857 + }, + { + "epoch": 5.477635782747604, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 6858 + }, + { + "epoch": 5.478434504792332, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 6859 + }, + { + "epoch": 5.479233226837061, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 6860 + }, + { + "epoch": 5.4800319488817895, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 6861 + }, + { + "epoch": 5.480830670926517, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 6862 + }, + { + "epoch": 5.481629392971246, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 6863 + }, + { + "epoch": 5.482428115015974, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 6864 + }, + { + "epoch": 5.483226837060703, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 6865 + }, + { + "epoch": 5.484025559105431, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 6866 + }, + { + "epoch": 5.48482428115016, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 6867 + }, + { + "epoch": 5.485623003194888, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 6868 + }, + { + "epoch": 5.486421725239617, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 6869 + }, + { + "epoch": 5.487220447284345, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 6870 + }, + { + "epoch": 5.488019169329074, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 6871 + }, + { + "epoch": 5.488817891373802, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 6872 + }, + { + "epoch": 5.48961661341853, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 6873 + }, + { + "epoch": 5.4904153354632586, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 6874 + }, + { + "epoch": 5.491214057507987, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 6875 + }, + { + "epoch": 5.492012779552716, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 6876 + }, + { + "epoch": 5.492811501597444, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 6877 + }, + { + "epoch": 5.493610223642173, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 6878 + }, + { + "epoch": 5.494408945686901, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6879 + }, + { + "epoch": 5.49520766773163, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 6880 + }, + { + "epoch": 5.496006389776358, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 6881 + }, + { + "epoch": 5.496805111821086, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 6882 + }, + { + "epoch": 5.497603833865814, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 6883 + }, + { + "epoch": 5.498402555910543, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 6884 + }, + { + "epoch": 5.4992012779552715, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 6885 + }, + { + "epoch": 5.5, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 6886 + }, + { + "epoch": 5.5007987220447285, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 6887 + }, + { + "epoch": 5.501597444089457, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 6888 + }, + { + "epoch": 5.502396166134186, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6889 + }, + { + "epoch": 5.503194888178914, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 6890 + }, + { + "epoch": 5.503993610223642, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 6891 + }, + { + "epoch": 5.50479233226837, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 6892 + }, + { + "epoch": 5.505591054313099, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 6893 + }, + { + "epoch": 5.506389776357827, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 6894 + }, + { + "epoch": 5.507188498402556, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 6895 + }, + { + "epoch": 5.507987220447284, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 6896 + }, + { + "epoch": 5.508785942492013, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 6897 + }, + { + "epoch": 5.5095846645367414, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 6898 + }, + { + "epoch": 5.51038338658147, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1191, + "step": 6899 + }, + { + "epoch": 5.511182108626198, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 6900 + }, + { + "epoch": 5.511980830670926, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 6901 + }, + { + "epoch": 5.512779552715655, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 6902 + }, + { + "epoch": 5.513578274760383, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 6903 + }, + { + "epoch": 5.514376996805112, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 6904 + }, + { + "epoch": 5.51517571884984, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 6905 + }, + { + "epoch": 5.515974440894569, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 6906 + }, + { + "epoch": 5.516773162939297, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 6907 + }, + { + "epoch": 5.517571884984026, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 6908 + }, + { + "epoch": 5.518370607028754, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 6909 + }, + { + "epoch": 5.519169329073483, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 6910 + }, + { + "epoch": 5.5199680511182105, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 6911 + }, + { + "epoch": 5.520766773162939, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 6912 + }, + { + "epoch": 5.521565495207668, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 6913 + }, + { + "epoch": 5.522364217252396, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 6914 + }, + { + "epoch": 5.523162939297125, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 6915 + }, + { + "epoch": 5.523961661341853, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 6916 + }, + { + "epoch": 5.524760383386582, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 6917 + }, + { + "epoch": 5.52555910543131, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 6918 + }, + { + "epoch": 5.526357827476039, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 6919 + }, + { + "epoch": 5.527156549520766, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 6920 + }, + { + "epoch": 5.527955271565495, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 6921 + }, + { + "epoch": 5.5287539936102235, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 6922 + }, + { + "epoch": 5.529552715654952, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6923 + }, + { + "epoch": 5.5303514376996805, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 6924 + }, + { + "epoch": 5.531150159744409, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1228, + "step": 6925 + }, + { + "epoch": 5.531948881789138, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 6926 + }, + { + "epoch": 5.532747603833866, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 6927 + }, + { + "epoch": 5.533546325878595, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 6928 + }, + { + "epoch": 5.534345047923322, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 6929 + }, + { + "epoch": 5.535143769968051, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 6930 + }, + { + "epoch": 5.535942492012779, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.121, + "step": 6931 + }, + { + "epoch": 5.536741214057508, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 6932 + }, + { + "epoch": 5.537539936102236, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 6933 + }, + { + "epoch": 5.538338658146965, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 6934 + }, + { + "epoch": 5.539137380191693, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 6935 + }, + { + "epoch": 5.539936102236422, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 6936 + }, + { + "epoch": 5.5407348242811505, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 6937 + }, + { + "epoch": 5.541533546325878, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 6938 + }, + { + "epoch": 5.542332268370607, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 6939 + }, + { + "epoch": 5.543130990415335, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 6940 + }, + { + "epoch": 5.543929712460064, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 6941 + }, + { + "epoch": 5.544728434504792, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1216, + "step": 6942 + }, + { + "epoch": 5.545527156549521, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 6943 + }, + { + "epoch": 5.546325878594249, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 6944 + }, + { + "epoch": 5.547124600638978, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 6945 + }, + { + "epoch": 5.547923322683706, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 6946 + }, + { + "epoch": 5.548722044728435, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 6947 + }, + { + "epoch": 5.549520766773163, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 6948 + }, + { + "epoch": 5.550319488817891, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 6949 + }, + { + "epoch": 5.55111821086262, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 6950 + }, + { + "epoch": 5.551916932907348, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 6951 + }, + { + "epoch": 5.552715654952077, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 6952 + }, + { + "epoch": 5.553514376996805, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 6953 + }, + { + "epoch": 5.554313099041534, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 6954 + }, + { + "epoch": 5.555111821086262, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 6955 + }, + { + "epoch": 5.555910543130991, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 6956 + }, + { + "epoch": 5.556709265175719, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 6957 + }, + { + "epoch": 5.557507987220447, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 6958 + }, + { + "epoch": 5.5583067092651754, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 6959 + }, + { + "epoch": 5.559105431309904, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 6960 + }, + { + "epoch": 5.5599041533546325, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 6961 + }, + { + "epoch": 5.560702875399361, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 6962 + }, + { + "epoch": 5.56150159744409, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 6963 + }, + { + "epoch": 5.562300319488818, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 6964 + }, + { + "epoch": 5.563099041533547, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 6965 + }, + { + "epoch": 5.563897763578275, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 6966 + }, + { + "epoch": 5.564696485623003, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 6967 + }, + { + "epoch": 5.565495207667731, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 6968 + }, + { + "epoch": 5.56629392971246, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 6969 + }, + { + "epoch": 5.567092651757188, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 6970 + }, + { + "epoch": 5.567891373801917, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 6971 + }, + { + "epoch": 5.568690095846645, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 6972 + }, + { + "epoch": 5.569488817891374, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 6973 + }, + { + "epoch": 5.5702875399361025, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 6974 + }, + { + "epoch": 5.571086261980831, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 6975 + }, + { + "epoch": 5.571884984025559, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1168, + "step": 6976 + }, + { + "epoch": 5.572683706070287, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 6977 + }, + { + "epoch": 5.573482428115016, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 6978 + }, + { + "epoch": 5.574281150159744, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 6979 + }, + { + "epoch": 5.575079872204473, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 6980 + }, + { + "epoch": 5.575878594249201, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 6981 + }, + { + "epoch": 5.57667731629393, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 6982 + }, + { + "epoch": 5.577476038338658, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 6983 + }, + { + "epoch": 5.578274760383387, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 6984 + }, + { + "epoch": 5.5790734824281145, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 6985 + }, + { + "epoch": 5.579872204472844, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 6986 + }, + { + "epoch": 5.580670926517572, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1224, + "step": 6987 + }, + { + "epoch": 5.5814696485623, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 6988 + }, + { + "epoch": 5.582268370607029, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 6989 + }, + { + "epoch": 5.583067092651757, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 6990 + }, + { + "epoch": 5.583865814696486, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 6991 + }, + { + "epoch": 5.584664536741214, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 6992 + }, + { + "epoch": 5.585463258785943, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 6993 + }, + { + "epoch": 5.586261980830671, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 6994 + }, + { + "epoch": 5.5870607028754, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 6995 + }, + { + "epoch": 5.587859424920127, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 6996 + }, + { + "epoch": 5.588658146964856, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 6997 + }, + { + "epoch": 5.5894568690095845, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 6998 + }, + { + "epoch": 5.590255591054313, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 6999 + }, + { + "epoch": 5.5910543130990416, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 7000 + }, + { + "epoch": 5.59185303514377, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 7001 + }, + { + "epoch": 5.592651757188499, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 7002 + }, + { + "epoch": 5.593450479233227, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 7003 + }, + { + "epoch": 5.594249201277956, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 7004 + }, + { + "epoch": 5.595047923322683, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 7005 + }, + { + "epoch": 5.595846645367412, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 7006 + }, + { + "epoch": 5.59664536741214, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 7007 + }, + { + "epoch": 5.597444089456869, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 7008 + }, + { + "epoch": 5.598242811501597, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 7009 + }, + { + "epoch": 5.599041533546326, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 7010 + }, + { + "epoch": 5.5998402555910545, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1233, + "step": 7011 + }, + { + "epoch": 5.600638977635783, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 7012 + }, + { + "epoch": 5.6014376996805115, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 7013 + }, + { + "epoch": 5.602236421725239, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 7014 + }, + { + "epoch": 5.603035143769968, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 7015 + }, + { + "epoch": 5.603833865814696, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 7016 + }, + { + "epoch": 5.604632587859425, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 7017 + }, + { + "epoch": 5.605431309904153, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 7018 + }, + { + "epoch": 5.606230031948882, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 7019 + }, + { + "epoch": 5.60702875399361, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 7020 + }, + { + "epoch": 5.607827476038339, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 7021 + }, + { + "epoch": 5.608626198083067, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 7022 + }, + { + "epoch": 5.609424920127795, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 7023 + }, + { + "epoch": 5.6102236421725244, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 7024 + }, + { + "epoch": 5.611022364217252, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 7025 + }, + { + "epoch": 5.611821086261981, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7026 + }, + { + "epoch": 5.612619808306709, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 7027 + }, + { + "epoch": 5.613418530351438, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 7028 + }, + { + "epoch": 5.614217252396166, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 7029 + }, + { + "epoch": 5.615015974440895, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 7030 + }, + { + "epoch": 5.615814696485623, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7031 + }, + { + "epoch": 5.616613418530352, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 7032 + }, + { + "epoch": 5.61741214057508, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 7033 + }, + { + "epoch": 5.618210862619808, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 7034 + }, + { + "epoch": 5.6190095846645365, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 7035 + }, + { + "epoch": 5.619808306709265, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 7036 + }, + { + "epoch": 5.6206070287539935, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 7037 + }, + { + "epoch": 5.621405750798722, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 7038 + }, + { + "epoch": 5.622204472843451, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 7039 + }, + { + "epoch": 5.623003194888179, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 7040 + }, + { + "epoch": 5.623801916932908, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 7041 + }, + { + "epoch": 5.624600638977636, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7042 + }, + { + "epoch": 5.625399361022364, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 7043 + }, + { + "epoch": 5.626198083067092, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 7044 + }, + { + "epoch": 5.626996805111821, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 7045 + }, + { + "epoch": 5.627795527156549, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 7046 + }, + { + "epoch": 5.628594249201278, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 7047 + }, + { + "epoch": 5.6293929712460065, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 7048 + }, + { + "epoch": 5.630191693290735, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 7049 + }, + { + "epoch": 5.6309904153354635, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 7050 + }, + { + "epoch": 5.631789137380192, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 7051 + }, + { + "epoch": 5.63258785942492, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 7052 + }, + { + "epoch": 5.633386581469648, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 7053 + }, + { + "epoch": 5.634185303514377, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 7054 + }, + { + "epoch": 5.634984025559105, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 7055 + }, + { + "epoch": 5.635782747603834, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 7056 + }, + { + "epoch": 5.636581469648562, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 7057 + }, + { + "epoch": 5.637380191693291, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 7058 + }, + { + "epoch": 5.638178913738019, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 7059 + }, + { + "epoch": 5.638977635782748, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 7060 + }, + { + "epoch": 5.6397763578274756, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 7061 + }, + { + "epoch": 5.640575079872205, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 7062 + }, + { + "epoch": 5.641373801916933, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 7063 + }, + { + "epoch": 5.642172523961661, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 7064 + }, + { + "epoch": 5.64297124600639, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7065 + }, + { + "epoch": 5.643769968051118, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 7066 + }, + { + "epoch": 5.644568690095847, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 7067 + }, + { + "epoch": 5.645367412140575, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 7068 + }, + { + "epoch": 5.646166134185304, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 7069 + }, + { + "epoch": 5.646964856230032, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 7070 + }, + { + "epoch": 5.647763578274761, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 7071 + }, + { + "epoch": 5.6485623003194885, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 7072 + }, + { + "epoch": 5.649361022364217, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 7073 + }, + { + "epoch": 5.6501597444089455, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 7074 + }, + { + "epoch": 5.650958466453674, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 7075 + }, + { + "epoch": 5.651757188498403, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 7076 + }, + { + "epoch": 5.652555910543131, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7077 + }, + { + "epoch": 5.65335463258786, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 7078 + }, + { + "epoch": 5.654153354632588, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 7079 + }, + { + "epoch": 5.654952076677317, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 7080 + }, + { + "epoch": 5.655750798722044, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 7081 + }, + { + "epoch": 5.656549520766773, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 7082 + }, + { + "epoch": 5.657348242811501, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 7083 + }, + { + "epoch": 5.65814696485623, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 7084 + }, + { + "epoch": 5.6589456869009584, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 7085 + }, + { + "epoch": 5.659744408945687, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 7086 + }, + { + "epoch": 5.6605431309904155, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 7087 + }, + { + "epoch": 5.661341853035144, + "grad_norm": 0.36328125, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 7088 + }, + { + "epoch": 5.662140575079873, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 7089 + }, + { + "epoch": 5.6629392971246, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7090 + }, + { + "epoch": 5.663738019169329, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 7091 + }, + { + "epoch": 5.664536741214057, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7092 + }, + { + "epoch": 5.665335463258786, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 7093 + }, + { + "epoch": 5.666134185303514, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 7094 + }, + { + "epoch": 5.666932907348243, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 7095 + }, + { + "epoch": 5.667731629392971, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7096 + }, + { + "epoch": 5.6685303514377, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 7097 + }, + { + "epoch": 5.669329073482428, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 7098 + }, + { + "epoch": 5.670127795527156, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 7099 + }, + { + "epoch": 5.6709265175718855, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 7100 + }, + { + "epoch": 5.671725239616613, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 7101 + }, + { + "epoch": 5.672523961661342, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 7102 + }, + { + "epoch": 5.67332268370607, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 7103 + }, + { + "epoch": 5.674121405750799, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 7104 + }, + { + "epoch": 5.674920127795527, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 7105 + }, + { + "epoch": 5.675718849840256, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 7106 + }, + { + "epoch": 5.676517571884984, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 7107 + }, + { + "epoch": 5.677316293929713, + "grad_norm": 0.345703125, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 7108 + }, + { + "epoch": 5.678115015974441, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 7109 + }, + { + "epoch": 5.678913738019169, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 7110 + }, + { + "epoch": 5.6797124600638975, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 7111 + }, + { + "epoch": 5.680511182108626, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 7112 + }, + { + "epoch": 5.681309904153355, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 7113 + }, + { + "epoch": 5.682108626198083, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 7114 + }, + { + "epoch": 5.682907348242812, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 7115 + }, + { + "epoch": 5.68370607028754, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 7116 + }, + { + "epoch": 5.684504792332269, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 7117 + }, + { + "epoch": 5.685303514376997, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 7118 + }, + { + "epoch": 5.686102236421725, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 7119 + }, + { + "epoch": 5.686900958466453, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 7120 + }, + { + "epoch": 5.687699680511182, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 7121 + }, + { + "epoch": 5.68849840255591, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 7122 + }, + { + "epoch": 5.689297124600639, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 7123 + }, + { + "epoch": 5.6900958466453675, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 7124 + }, + { + "epoch": 5.690894568690096, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 7125 + }, + { + "epoch": 5.6916932907348246, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 7126 + }, + { + "epoch": 5.692492012779553, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 7127 + }, + { + "epoch": 5.693290734824281, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 7128 + }, + { + "epoch": 5.694089456869009, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 7129 + }, + { + "epoch": 5.694888178913738, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7130 + }, + { + "epoch": 5.695686900958466, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 7131 + }, + { + "epoch": 5.696485623003195, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 7132 + }, + { + "epoch": 5.697284345047923, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7133 + }, + { + "epoch": 5.698083067092652, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 7134 + }, + { + "epoch": 5.69888178913738, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 7135 + }, + { + "epoch": 5.699680511182109, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 7136 + }, + { + "epoch": 5.700479233226837, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 7137 + }, + { + "epoch": 5.701277955271565, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 7138 + }, + { + "epoch": 5.702076677316294, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 7139 + }, + { + "epoch": 5.702875399361022, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 7140 + }, + { + "epoch": 5.703674121405751, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 7141 + }, + { + "epoch": 5.704472843450479, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 7142 + }, + { + "epoch": 5.705271565495208, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 7143 + }, + { + "epoch": 5.706070287539936, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 7144 + }, + { + "epoch": 5.706869009584665, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 7145 + }, + { + "epoch": 5.707667731629393, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7146 + }, + { + "epoch": 5.708466453674122, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 7147 + }, + { + "epoch": 5.7092651757188495, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7148 + }, + { + "epoch": 5.710063897763578, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7149 + }, + { + "epoch": 5.710862619808307, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 7150 + }, + { + "epoch": 5.711661341853035, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 7151 + }, + { + "epoch": 5.712460063897764, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 7152 + }, + { + "epoch": 5.713258785942492, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 7153 + }, + { + "epoch": 5.714057507987221, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 7154 + }, + { + "epoch": 5.714856230031949, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 7155 + }, + { + "epoch": 5.715654952076678, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 7156 + }, + { + "epoch": 5.716453674121405, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 7157 + }, + { + "epoch": 5.717252396166134, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 7158 + }, + { + "epoch": 5.718051118210862, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 7159 + }, + { + "epoch": 5.718849840255591, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 7160 + }, + { + "epoch": 5.7196485623003195, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 7161 + }, + { + "epoch": 5.720447284345048, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 7162 + }, + { + "epoch": 5.7212460063897765, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 7163 + }, + { + "epoch": 5.722044728434505, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 7164 + }, + { + "epoch": 5.722843450479234, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 7165 + }, + { + "epoch": 5.723642172523961, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 7166 + }, + { + "epoch": 5.72444089456869, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 7167 + }, + { + "epoch": 5.725239616613418, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 7168 + }, + { + "epoch": 5.726038338658147, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 7169 + }, + { + "epoch": 5.726837060702875, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 7170 + }, + { + "epoch": 5.727635782747604, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 7171 + }, + { + "epoch": 5.728434504792332, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 7172 + }, + { + "epoch": 5.729233226837061, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 7173 + }, + { + "epoch": 5.7300319488817895, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7174 + }, + { + "epoch": 5.730830670926517, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 7175 + }, + { + "epoch": 5.731629392971246, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 7176 + }, + { + "epoch": 5.732428115015974, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 7177 + }, + { + "epoch": 5.733226837060703, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 7178 + }, + { + "epoch": 5.734025559105431, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 7179 + }, + { + "epoch": 5.73482428115016, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 7180 + }, + { + "epoch": 5.735623003194888, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 7181 + }, + { + "epoch": 5.736421725239617, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 7182 + }, + { + "epoch": 5.737220447284345, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 7183 + }, + { + "epoch": 5.738019169329074, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 7184 + }, + { + "epoch": 5.738817891373802, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 7185 + }, + { + "epoch": 5.73961661341853, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 7186 + }, + { + "epoch": 5.7404153354632586, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 7187 + }, + { + "epoch": 5.741214057507987, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 7188 + }, + { + "epoch": 5.742012779552716, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 7189 + }, + { + "epoch": 5.742811501597444, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 7190 + }, + { + "epoch": 5.743610223642173, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 7191 + }, + { + "epoch": 5.744408945686901, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 7192 + }, + { + "epoch": 5.74520766773163, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 7193 + }, + { + "epoch": 5.746006389776358, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 7194 + }, + { + "epoch": 5.746805111821086, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 7195 + }, + { + "epoch": 5.747603833865814, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 7196 + }, + { + "epoch": 5.748402555910543, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 7197 + }, + { + "epoch": 5.7492012779552715, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 7198 + }, + { + "epoch": 5.75, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 7199 + }, + { + "epoch": 5.7507987220447285, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 7200 + }, + { + "epoch": 5.751597444089457, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 7201 + }, + { + "epoch": 5.752396166134186, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 7202 + }, + { + "epoch": 5.753194888178914, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 7203 + }, + { + "epoch": 5.753993610223642, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 7204 + }, + { + "epoch": 5.75479233226837, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 7205 + }, + { + "epoch": 5.755591054313099, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 7206 + }, + { + "epoch": 5.756389776357827, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 7207 + }, + { + "epoch": 5.757188498402556, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 7208 + }, + { + "epoch": 5.757987220447284, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 7209 + }, + { + "epoch": 5.758785942492013, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 7210 + }, + { + "epoch": 5.7595846645367414, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 7211 + }, + { + "epoch": 5.76038338658147, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 7212 + }, + { + "epoch": 5.761182108626198, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 7213 + }, + { + "epoch": 5.761980830670926, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 7214 + }, + { + "epoch": 5.762779552715655, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 7215 + }, + { + "epoch": 5.763578274760383, + "grad_norm": 0.09619140625, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 7216 + }, + { + "epoch": 5.764376996805112, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 7217 + }, + { + "epoch": 5.76517571884984, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 7218 + }, + { + "epoch": 5.765974440894569, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 7219 + }, + { + "epoch": 5.766773162939297, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 7220 + }, + { + "epoch": 5.767571884984026, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 7221 + }, + { + "epoch": 5.768370607028754, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 7222 + }, + { + "epoch": 5.769169329073483, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 7223 + }, + { + "epoch": 5.7699680511182105, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 7224 + }, + { + "epoch": 5.770766773162939, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 7225 + }, + { + "epoch": 5.771565495207668, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 7226 + }, + { + "epoch": 5.772364217252396, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 7227 + }, + { + "epoch": 5.773162939297125, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 7228 + }, + { + "epoch": 5.773961661341853, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 7229 + }, + { + "epoch": 5.774760383386582, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 7230 + }, + { + "epoch": 5.77555910543131, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 7231 + }, + { + "epoch": 5.776357827476039, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7232 + }, + { + "epoch": 5.777156549520766, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 7233 + }, + { + "epoch": 5.777955271565495, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 7234 + }, + { + "epoch": 5.7787539936102235, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 7235 + }, + { + "epoch": 5.779552715654952, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 7236 + }, + { + "epoch": 5.7803514376996805, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 7237 + }, + { + "epoch": 5.781150159744409, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 7238 + }, + { + "epoch": 5.781948881789138, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 7239 + }, + { + "epoch": 5.782747603833866, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 7240 + }, + { + "epoch": 5.783546325878595, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 7241 + }, + { + "epoch": 5.784345047923322, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 7242 + }, + { + "epoch": 5.785143769968051, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 7243 + }, + { + "epoch": 5.785942492012779, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 7244 + }, + { + "epoch": 5.786741214057508, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 7245 + }, + { + "epoch": 5.787539936102236, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 7246 + }, + { + "epoch": 5.788338658146965, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 7247 + }, + { + "epoch": 5.789137380191693, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 7248 + }, + { + "epoch": 5.789936102236422, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 7249 + }, + { + "epoch": 5.7907348242811505, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 7250 + }, + { + "epoch": 5.791533546325878, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 7251 + }, + { + "epoch": 5.792332268370607, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 7252 + }, + { + "epoch": 5.793130990415335, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 7253 + }, + { + "epoch": 5.793929712460064, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 7254 + }, + { + "epoch": 5.794728434504792, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 7255 + }, + { + "epoch": 5.795527156549521, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 7256 + }, + { + "epoch": 5.796325878594249, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 7257 + }, + { + "epoch": 5.797124600638978, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 7258 + }, + { + "epoch": 5.797923322683706, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 7259 + }, + { + "epoch": 5.798722044728435, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 7260 + }, + { + "epoch": 5.799520766773163, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 7261 + }, + { + "epoch": 5.800319488817891, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 7262 + }, + { + "epoch": 5.80111821086262, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 7263 + }, + { + "epoch": 5.801916932907348, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 7264 + }, + { + "epoch": 5.802715654952077, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 7265 + }, + { + "epoch": 5.803514376996805, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 7266 + }, + { + "epoch": 5.804313099041534, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 7267 + }, + { + "epoch": 5.805111821086262, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 7268 + }, + { + "epoch": 5.805910543130991, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 7269 + }, + { + "epoch": 5.806709265175719, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 7270 + }, + { + "epoch": 5.807507987220447, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 7271 + }, + { + "epoch": 5.8083067092651754, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 7272 + }, + { + "epoch": 5.809105431309904, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 7273 + }, + { + "epoch": 5.8099041533546325, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 7274 + }, + { + "epoch": 5.810702875399361, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 7275 + }, + { + "epoch": 5.81150159744409, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7276 + }, + { + "epoch": 5.812300319488818, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 7277 + }, + { + "epoch": 5.813099041533547, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 7278 + }, + { + "epoch": 5.813897763578275, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 7279 + }, + { + "epoch": 5.814696485623003, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 7280 + }, + { + "epoch": 5.815495207667731, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 7281 + }, + { + "epoch": 5.81629392971246, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 7282 + }, + { + "epoch": 5.817092651757188, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 7283 + }, + { + "epoch": 5.817891373801917, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 7284 + }, + { + "epoch": 5.818690095846645, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 7285 + }, + { + "epoch": 5.819488817891374, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 7286 + }, + { + "epoch": 5.8202875399361025, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 7287 + }, + { + "epoch": 5.821086261980831, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 7288 + }, + { + "epoch": 5.821884984025559, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 7289 + }, + { + "epoch": 5.822683706070287, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 7290 + }, + { + "epoch": 5.823482428115016, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 7291 + }, + { + "epoch": 5.824281150159744, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 7292 + }, + { + "epoch": 5.825079872204473, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 7293 + }, + { + "epoch": 5.825878594249201, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 7294 + }, + { + "epoch": 5.82667731629393, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 7295 + }, + { + "epoch": 5.827476038338658, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 7296 + }, + { + "epoch": 5.828274760383387, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 7297 + }, + { + "epoch": 5.8290734824281145, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 7298 + }, + { + "epoch": 5.829872204472844, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 7299 + }, + { + "epoch": 5.830670926517572, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7300 + }, + { + "epoch": 5.8314696485623, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 7301 + }, + { + "epoch": 5.832268370607029, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 7302 + }, + { + "epoch": 5.833067092651757, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 7303 + }, + { + "epoch": 5.833865814696486, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 7304 + }, + { + "epoch": 5.834664536741214, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 7305 + }, + { + "epoch": 5.835463258785943, + "grad_norm": 0.384765625, + "learning_rate": 0.0001, + "loss": 1.1184, + "step": 7306 + }, + { + "epoch": 5.836261980830671, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 7307 + }, + { + "epoch": 5.8370607028754, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 7308 + }, + { + "epoch": 5.837859424920127, + "grad_norm": 0.392578125, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 7309 + }, + { + "epoch": 5.838658146964856, + "grad_norm": 0.365234375, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 7310 + }, + { + "epoch": 5.8394568690095845, + "grad_norm": 0.35546875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 7311 + }, + { + "epoch": 5.840255591054313, + "grad_norm": 0.54296875, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 7312 + }, + { + "epoch": 5.8410543130990416, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 7313 + }, + { + "epoch": 5.84185303514377, + "grad_norm": 0.578125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 7314 + }, + { + "epoch": 5.842651757188499, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 7315 + }, + { + "epoch": 5.843450479233227, + "grad_norm": 0.59765625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 7316 + }, + { + "epoch": 5.844249201277956, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 7317 + }, + { + "epoch": 5.845047923322683, + "grad_norm": 0.59375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 7318 + }, + { + "epoch": 5.845846645367412, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7319 + }, + { + "epoch": 5.84664536741214, + "grad_norm": 0.54296875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 7320 + }, + { + "epoch": 5.847444089456869, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 7321 + }, + { + "epoch": 5.848242811501597, + "grad_norm": 0.48046875, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 7322 + }, + { + "epoch": 5.849041533546326, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7323 + }, + { + "epoch": 5.8498402555910545, + "grad_norm": 0.4609375, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 7324 + }, + { + "epoch": 5.850638977635783, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 7325 + }, + { + "epoch": 5.8514376996805115, + "grad_norm": 0.6484375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 7326 + }, + { + "epoch": 5.852236421725239, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 7327 + }, + { + "epoch": 5.853035143769968, + "grad_norm": 0.640625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 7328 + }, + { + "epoch": 5.853833865814696, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 7329 + }, + { + "epoch": 5.854632587859425, + "grad_norm": 0.75, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 7330 + }, + { + "epoch": 5.855431309904153, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 7331 + }, + { + "epoch": 5.856230031948882, + "grad_norm": 0.765625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 7332 + }, + { + "epoch": 5.85702875399361, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 7333 + }, + { + "epoch": 5.857827476038339, + "grad_norm": 0.640625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 7334 + }, + { + "epoch": 5.858626198083067, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 7335 + }, + { + "epoch": 5.859424920127795, + "grad_norm": 0.66796875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 7336 + }, + { + "epoch": 5.8602236421725244, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 7337 + }, + { + "epoch": 5.861022364217252, + "grad_norm": 0.578125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 7338 + }, + { + "epoch": 5.861821086261981, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 7339 + }, + { + "epoch": 5.862619808306709, + "grad_norm": 0.384765625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 7340 + }, + { + "epoch": 5.863418530351438, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 7341 + }, + { + "epoch": 5.864217252396166, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1198, + "step": 7342 + }, + { + "epoch": 5.865015974440895, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 7343 + }, + { + "epoch": 5.865814696485623, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 7344 + }, + { + "epoch": 5.866613418530352, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 7345 + }, + { + "epoch": 5.86741214057508, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 7346 + }, + { + "epoch": 5.868210862619808, + "grad_norm": 0.328125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 7347 + }, + { + "epoch": 5.8690095846645365, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 7348 + }, + { + "epoch": 5.869808306709265, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 7349 + }, + { + "epoch": 5.8706070287539935, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 7350 + }, + { + "epoch": 5.871405750798722, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 7351 + }, + { + "epoch": 5.872204472843451, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 7352 + }, + { + "epoch": 5.873003194888179, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 7353 + }, + { + "epoch": 5.873801916932908, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 7354 + }, + { + "epoch": 5.874600638977636, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 7355 + }, + { + "epoch": 5.875399361022364, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 7356 + }, + { + "epoch": 5.876198083067092, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 7357 + }, + { + "epoch": 5.876996805111821, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 7358 + }, + { + "epoch": 5.877795527156549, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 7359 + }, + { + "epoch": 5.878594249201278, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 7360 + }, + { + "epoch": 5.8793929712460065, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 7361 + }, + { + "epoch": 5.880191693290735, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 7362 + }, + { + "epoch": 5.8809904153354635, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 7363 + }, + { + "epoch": 5.881789137380192, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 7364 + }, + { + "epoch": 5.88258785942492, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 7365 + }, + { + "epoch": 5.883386581469648, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7366 + }, + { + "epoch": 5.884185303514377, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 7367 + }, + { + "epoch": 5.884984025559105, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 7368 + }, + { + "epoch": 5.885782747603834, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 7369 + }, + { + "epoch": 5.886581469648562, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 7370 + }, + { + "epoch": 5.887380191693291, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 7371 + }, + { + "epoch": 5.888178913738019, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 7372 + }, + { + "epoch": 5.888977635782748, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 7373 + }, + { + "epoch": 5.8897763578274756, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 7374 + }, + { + "epoch": 5.890575079872205, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7375 + }, + { + "epoch": 5.891373801916933, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 7376 + }, + { + "epoch": 5.892172523961661, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 7377 + }, + { + "epoch": 5.89297124600639, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 7378 + }, + { + "epoch": 5.893769968051118, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 7379 + }, + { + "epoch": 5.894568690095847, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 7380 + }, + { + "epoch": 5.895367412140575, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 7381 + }, + { + "epoch": 5.896166134185304, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 7382 + }, + { + "epoch": 5.896964856230032, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 7383 + }, + { + "epoch": 5.897763578274761, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 7384 + }, + { + "epoch": 5.8985623003194885, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 7385 + }, + { + "epoch": 5.899361022364217, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 7386 + }, + { + "epoch": 5.9001597444089455, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 7387 + }, + { + "epoch": 5.900958466453674, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 7388 + }, + { + "epoch": 5.901757188498403, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 7389 + }, + { + "epoch": 5.902555910543131, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 7390 + }, + { + "epoch": 5.90335463258786, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.12, + "step": 7391 + }, + { + "epoch": 5.904153354632588, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 7392 + }, + { + "epoch": 5.904952076677317, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 7393 + }, + { + "epoch": 5.905750798722044, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 7394 + }, + { + "epoch": 5.906549520766773, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 7395 + }, + { + "epoch": 5.907348242811501, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 7396 + }, + { + "epoch": 5.90814696485623, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 7397 + }, + { + "epoch": 5.9089456869009584, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 7398 + }, + { + "epoch": 5.909744408945687, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 7399 + }, + { + "epoch": 5.9105431309904155, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1217, + "step": 7400 + }, + { + "epoch": 5.911341853035144, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 7401 + }, + { + "epoch": 5.912140575079873, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 7402 + }, + { + "epoch": 5.9129392971246, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 7403 + }, + { + "epoch": 5.913738019169329, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 7404 + }, + { + "epoch": 5.914536741214057, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 7405 + }, + { + "epoch": 5.915335463258786, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 7406 + }, + { + "epoch": 5.916134185303514, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 7407 + }, + { + "epoch": 5.916932907348243, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 7408 + }, + { + "epoch": 5.917731629392971, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 7409 + }, + { + "epoch": 5.9185303514377, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 7410 + }, + { + "epoch": 5.919329073482428, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7411 + }, + { + "epoch": 5.920127795527156, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 7412 + }, + { + "epoch": 5.9209265175718855, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 7413 + }, + { + "epoch": 5.921725239616613, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 7414 + }, + { + "epoch": 5.922523961661342, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 7415 + }, + { + "epoch": 5.92332268370607, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 7416 + }, + { + "epoch": 5.924121405750799, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 7417 + }, + { + "epoch": 5.924920127795527, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 7418 + }, + { + "epoch": 5.925718849840256, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 7419 + }, + { + "epoch": 5.926517571884984, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 7420 + }, + { + "epoch": 5.927316293929713, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 7421 + }, + { + "epoch": 5.928115015974441, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 7422 + }, + { + "epoch": 5.928913738019169, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 7423 + }, + { + "epoch": 5.9297124600638975, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 7424 + }, + { + "epoch": 5.930511182108626, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 7425 + }, + { + "epoch": 5.931309904153355, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 7426 + }, + { + "epoch": 5.932108626198083, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 7427 + }, + { + "epoch": 5.932907348242812, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 7428 + }, + { + "epoch": 5.93370607028754, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 7429 + }, + { + "epoch": 5.934504792332269, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 7430 + }, + { + "epoch": 5.935303514376997, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 7431 + }, + { + "epoch": 5.936102236421725, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 7432 + }, + { + "epoch": 5.936900958466453, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 7433 + }, + { + "epoch": 5.937699680511182, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 7434 + }, + { + "epoch": 5.93849840255591, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 7435 + }, + { + "epoch": 5.939297124600639, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 7436 + }, + { + "epoch": 5.9400958466453675, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 7437 + }, + { + "epoch": 5.940894568690096, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7438 + }, + { + "epoch": 5.9416932907348246, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 7439 + }, + { + "epoch": 5.942492012779553, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 7440 + }, + { + "epoch": 5.943290734824281, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 7441 + }, + { + "epoch": 5.944089456869009, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1241, + "step": 7442 + }, + { + "epoch": 5.944888178913738, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 7443 + }, + { + "epoch": 5.945686900958466, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 7444 + }, + { + "epoch": 5.946485623003195, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 7445 + }, + { + "epoch": 5.947284345047923, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 7446 + }, + { + "epoch": 5.948083067092652, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 7447 + }, + { + "epoch": 5.94888178913738, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 7448 + }, + { + "epoch": 5.949680511182109, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 7449 + }, + { + "epoch": 5.950479233226837, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 7450 + }, + { + "epoch": 5.951277955271565, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 7451 + }, + { + "epoch": 5.952076677316294, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 7452 + }, + { + "epoch": 5.952875399361022, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7453 + }, + { + "epoch": 5.953674121405751, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 7454 + }, + { + "epoch": 5.954472843450479, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1209, + "step": 7455 + }, + { + "epoch": 5.955271565495208, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 7456 + }, + { + "epoch": 5.956070287539936, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 7457 + }, + { + "epoch": 5.956869009584665, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 7458 + }, + { + "epoch": 5.957667731629393, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 7459 + }, + { + "epoch": 5.958466453674122, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 7460 + }, + { + "epoch": 5.9592651757188495, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 7461 + }, + { + "epoch": 5.960063897763578, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 7462 + }, + { + "epoch": 5.960862619808307, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 7463 + }, + { + "epoch": 5.961661341853035, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 7464 + }, + { + "epoch": 5.962460063897764, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 7465 + }, + { + "epoch": 5.963258785942492, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 7466 + }, + { + "epoch": 5.964057507987221, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 7467 + }, + { + "epoch": 5.964856230031949, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 7468 + }, + { + "epoch": 5.965654952076678, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 7469 + }, + { + "epoch": 5.966453674121405, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 7470 + }, + { + "epoch": 5.967252396166134, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 7471 + }, + { + "epoch": 5.968051118210862, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 7472 + }, + { + "epoch": 5.968849840255591, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 7473 + }, + { + "epoch": 5.9696485623003195, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 7474 + }, + { + "epoch": 5.970447284345048, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 7475 + }, + { + "epoch": 5.9712460063897765, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 7476 + }, + { + "epoch": 5.972044728434505, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 7477 + }, + { + "epoch": 5.972843450479234, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 7478 + }, + { + "epoch": 5.973642172523961, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7479 + }, + { + "epoch": 5.97444089456869, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 7480 + }, + { + "epoch": 5.975239616613418, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 7481 + }, + { + "epoch": 5.976038338658147, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 7482 + }, + { + "epoch": 5.976837060702875, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 7483 + }, + { + "epoch": 5.977635782747604, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 7484 + }, + { + "epoch": 5.978434504792332, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 7485 + }, + { + "epoch": 5.979233226837061, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 7486 + }, + { + "epoch": 5.9800319488817895, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 7487 + }, + { + "epoch": 5.980830670926517, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7488 + }, + { + "epoch": 5.981629392971246, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 7489 + }, + { + "epoch": 5.982428115015974, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 7490 + }, + { + "epoch": 5.983226837060703, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 7491 + }, + { + "epoch": 5.984025559105431, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 7492 + }, + { + "epoch": 5.98482428115016, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 7493 + }, + { + "epoch": 5.985623003194888, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7494 + }, + { + "epoch": 5.986421725239617, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 7495 + }, + { + "epoch": 5.987220447284345, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 7496 + }, + { + "epoch": 5.988019169329074, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 7497 + }, + { + "epoch": 5.988817891373802, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 7498 + }, + { + "epoch": 5.98961661341853, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 7499 + }, + { + "epoch": 5.9904153354632586, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 7500 + }, + { + "epoch": 5.991214057507987, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 7501 + }, + { + "epoch": 5.992012779552716, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 7502 + }, + { + "epoch": 5.992811501597444, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 7503 + }, + { + "epoch": 5.993610223642173, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 7504 + }, + { + "epoch": 5.994408945686901, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 7505 + }, + { + "epoch": 5.99520766773163, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 7506 + }, + { + "epoch": 5.996006389776358, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 7507 + }, + { + "epoch": 5.996805111821086, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 7508 + }, + { + "epoch": 5.997603833865814, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 7509 + }, + { + "epoch": 5.998402555910543, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7510 + }, + { + "epoch": 5.9992012779552715, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 7511 + }, + { + "epoch": 6.0, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 7512 + }, + { + "epoch": 6.0007987220447285, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 7513 + }, + { + "epoch": 6.001597444089457, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 7514 + }, + { + "epoch": 6.002396166134186, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 7515 + }, + { + "epoch": 6.003194888178914, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 7516 + }, + { + "epoch": 6.003993610223642, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 7517 + }, + { + "epoch": 6.00479233226837, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 7518 + }, + { + "epoch": 6.005591054313099, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 7519 + }, + { + "epoch": 6.006389776357827, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 7520 + }, + { + "epoch": 6.007188498402556, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 7521 + }, + { + "epoch": 6.007987220447284, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 7522 + }, + { + "epoch": 6.008785942492013, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 7523 + }, + { + "epoch": 6.0095846645367414, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 7524 + }, + { + "epoch": 6.01038338658147, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 7525 + }, + { + "epoch": 6.0111821086261985, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 7526 + }, + { + "epoch": 6.011980830670926, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 7527 + }, + { + "epoch": 6.012779552715655, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1172, + "step": 7528 + }, + { + "epoch": 6.013578274760383, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 7529 + }, + { + "epoch": 6.014376996805112, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 7530 + }, + { + "epoch": 6.01517571884984, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 7531 + }, + { + "epoch": 6.015974440894569, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 7532 + }, + { + "epoch": 6.016773162939297, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 7533 + }, + { + "epoch": 6.017571884984026, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 7534 + }, + { + "epoch": 6.018370607028754, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 7535 + }, + { + "epoch": 6.019169329073482, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 7536 + }, + { + "epoch": 6.0199680511182105, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 7537 + }, + { + "epoch": 6.020766773162939, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 7538 + }, + { + "epoch": 6.021565495207668, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 7539 + }, + { + "epoch": 6.022364217252396, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 7540 + }, + { + "epoch": 6.023162939297125, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 7541 + }, + { + "epoch": 6.023961661341853, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 7542 + }, + { + "epoch": 6.024760383386582, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 7543 + }, + { + "epoch": 6.02555910543131, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 7544 + }, + { + "epoch": 6.026357827476039, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 7545 + }, + { + "epoch": 6.027156549520766, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 7546 + }, + { + "epoch": 6.027955271565495, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 7547 + }, + { + "epoch": 6.0287539936102235, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 7548 + }, + { + "epoch": 6.029552715654952, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 7549 + }, + { + "epoch": 6.0303514376996805, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 7550 + }, + { + "epoch": 6.031150159744409, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 7551 + }, + { + "epoch": 6.031948881789138, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 7552 + }, + { + "epoch": 6.032747603833866, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 7553 + }, + { + "epoch": 6.033546325878595, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 7554 + }, + { + "epoch": 6.034345047923322, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 7555 + }, + { + "epoch": 6.035143769968051, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 7556 + }, + { + "epoch": 6.035942492012779, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 7557 + }, + { + "epoch": 6.036741214057508, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 7558 + }, + { + "epoch": 6.037539936102236, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 7559 + }, + { + "epoch": 6.038338658146965, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 7560 + }, + { + "epoch": 6.039137380191693, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 7561 + }, + { + "epoch": 6.039936102236422, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 7562 + }, + { + "epoch": 6.0407348242811505, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 7563 + }, + { + "epoch": 6.041533546325879, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 7564 + }, + { + "epoch": 6.042332268370607, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 7565 + }, + { + "epoch": 6.043130990415335, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 7566 + }, + { + "epoch": 6.043929712460064, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 7567 + }, + { + "epoch": 6.044728434504792, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 7568 + }, + { + "epoch": 6.045527156549521, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 7569 + }, + { + "epoch": 6.046325878594249, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 7570 + }, + { + "epoch": 6.047124600638978, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 7571 + }, + { + "epoch": 6.047923322683706, + "grad_norm": 0.091796875, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 7572 + }, + { + "epoch": 6.048722044728435, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 7573 + }, + { + "epoch": 6.0495207667731625, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 7574 + }, + { + "epoch": 6.050319488817891, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 7575 + }, + { + "epoch": 6.05111821086262, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 7576 + }, + { + "epoch": 6.051916932907348, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7577 + }, + { + "epoch": 6.052715654952077, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 7578 + }, + { + "epoch": 6.053514376996805, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 7579 + }, + { + "epoch": 6.054313099041534, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 7580 + }, + { + "epoch": 6.055111821086262, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 7581 + }, + { + "epoch": 6.055910543130991, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1211, + "step": 7582 + }, + { + "epoch": 6.056709265175719, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 7583 + }, + { + "epoch": 6.057507987220447, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 7584 + }, + { + "epoch": 6.0583067092651754, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 7585 + }, + { + "epoch": 6.059105431309904, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 7586 + }, + { + "epoch": 6.0599041533546325, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 7587 + }, + { + "epoch": 6.060702875399361, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 7588 + }, + { + "epoch": 6.06150159744409, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 7589 + }, + { + "epoch": 6.062300319488818, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 7590 + }, + { + "epoch": 6.063099041533547, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 7591 + }, + { + "epoch": 6.063897763578275, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 7592 + }, + { + "epoch": 6.064696485623003, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 7593 + }, + { + "epoch": 6.065495207667731, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 7594 + }, + { + "epoch": 6.06629392971246, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 7595 + }, + { + "epoch": 6.067092651757188, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 7596 + }, + { + "epoch": 6.067891373801917, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 7597 + }, + { + "epoch": 6.068690095846645, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 7598 + }, + { + "epoch": 6.069488817891374, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7599 + }, + { + "epoch": 6.0702875399361025, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 7600 + }, + { + "epoch": 6.071086261980831, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 7601 + }, + { + "epoch": 6.0718849840255595, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 7602 + }, + { + "epoch": 6.072683706070287, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 7603 + }, + { + "epoch": 6.073482428115016, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 7604 + }, + { + "epoch": 6.074281150159744, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 7605 + }, + { + "epoch": 6.075079872204473, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 7606 + }, + { + "epoch": 6.075878594249201, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 7607 + }, + { + "epoch": 6.07667731629393, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 7608 + }, + { + "epoch": 6.077476038338658, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 7609 + }, + { + "epoch": 6.078274760383387, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 7610 + }, + { + "epoch": 6.079073482428115, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 7611 + }, + { + "epoch": 6.079872204472843, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 7612 + }, + { + "epoch": 6.080670926517572, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 7613 + }, + { + "epoch": 6.0814696485623, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 7614 + }, + { + "epoch": 6.082268370607029, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 7615 + }, + { + "epoch": 6.083067092651757, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 7616 + }, + { + "epoch": 6.083865814696486, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 7617 + }, + { + "epoch": 6.084664536741214, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 7618 + }, + { + "epoch": 6.085463258785943, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 7619 + }, + { + "epoch": 6.086261980830671, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 7620 + }, + { + "epoch": 6.0870607028754, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 7621 + }, + { + "epoch": 6.087859424920127, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 7622 + }, + { + "epoch": 6.088658146964856, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 7623 + }, + { + "epoch": 6.0894568690095845, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 7624 + }, + { + "epoch": 6.090255591054313, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 7625 + }, + { + "epoch": 6.0910543130990416, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 7626 + }, + { + "epoch": 6.09185303514377, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 7627 + }, + { + "epoch": 6.092651757188499, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 7628 + }, + { + "epoch": 6.093450479233227, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 7629 + }, + { + "epoch": 6.094249201277956, + "grad_norm": 0.33984375, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 7630 + }, + { + "epoch": 6.095047923322683, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 7631 + }, + { + "epoch": 6.095846645367412, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 7632 + }, + { + "epoch": 6.09664536741214, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 7633 + }, + { + "epoch": 6.097444089456869, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 7634 + }, + { + "epoch": 6.098242811501597, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 7635 + }, + { + "epoch": 6.099041533546326, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 7636 + }, + { + "epoch": 6.0998402555910545, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 7637 + }, + { + "epoch": 6.100638977635783, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 7638 + }, + { + "epoch": 6.1014376996805115, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 7639 + }, + { + "epoch": 6.102236421725239, + "grad_norm": 0.41015625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 7640 + }, + { + "epoch": 6.103035143769968, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 7641 + }, + { + "epoch": 6.103833865814696, + "grad_norm": 0.33984375, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 7642 + }, + { + "epoch": 6.104632587859425, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 7643 + }, + { + "epoch": 6.105431309904153, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 7644 + }, + { + "epoch": 6.106230031948882, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 7645 + }, + { + "epoch": 6.10702875399361, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7646 + }, + { + "epoch": 6.107827476038339, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 7647 + }, + { + "epoch": 6.108626198083067, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 7648 + }, + { + "epoch": 6.109424920127796, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 7649 + }, + { + "epoch": 6.110223642172524, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 7650 + }, + { + "epoch": 6.111022364217252, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7651 + }, + { + "epoch": 6.111821086261981, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 7652 + }, + { + "epoch": 6.112619808306709, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 7653 + }, + { + "epoch": 6.113418530351438, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 7654 + }, + { + "epoch": 6.114217252396166, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 7655 + }, + { + "epoch": 6.115015974440895, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 7656 + }, + { + "epoch": 6.115814696485623, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 7657 + }, + { + "epoch": 6.116613418530352, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 7658 + }, + { + "epoch": 6.11741214057508, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 7659 + }, + { + "epoch": 6.118210862619808, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7660 + }, + { + "epoch": 6.1190095846645365, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 7661 + }, + { + "epoch": 6.119808306709265, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 7662 + }, + { + "epoch": 6.1206070287539935, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 7663 + }, + { + "epoch": 6.121405750798722, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 7664 + }, + { + "epoch": 6.122204472843451, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 7665 + }, + { + "epoch": 6.123003194888179, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 7666 + }, + { + "epoch": 6.123801916932908, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 7667 + }, + { + "epoch": 6.124600638977636, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 7668 + }, + { + "epoch": 6.125399361022364, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 7669 + }, + { + "epoch": 6.126198083067092, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 7670 + }, + { + "epoch": 6.126996805111821, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 7671 + }, + { + "epoch": 6.127795527156549, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 7672 + }, + { + "epoch": 6.128594249201278, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 7673 + }, + { + "epoch": 6.1293929712460065, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 7674 + }, + { + "epoch": 6.130191693290735, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 7675 + }, + { + "epoch": 6.1309904153354635, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 7676 + }, + { + "epoch": 6.131789137380192, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 7677 + }, + { + "epoch": 6.13258785942492, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7678 + }, + { + "epoch": 6.133386581469648, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7679 + }, + { + "epoch": 6.134185303514377, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 7680 + }, + { + "epoch": 6.134984025559105, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 7681 + }, + { + "epoch": 6.135782747603834, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 7682 + }, + { + "epoch": 6.136581469648562, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 7683 + }, + { + "epoch": 6.137380191693291, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 7684 + }, + { + "epoch": 6.138178913738019, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 7685 + }, + { + "epoch": 6.138977635782748, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 7686 + }, + { + "epoch": 6.139776357827476, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 7687 + }, + { + "epoch": 6.140575079872204, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 7688 + }, + { + "epoch": 6.141373801916933, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 7689 + }, + { + "epoch": 6.142172523961661, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 7690 + }, + { + "epoch": 6.14297124600639, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 7691 + }, + { + "epoch": 6.143769968051118, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 7692 + }, + { + "epoch": 6.144568690095847, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 7693 + }, + { + "epoch": 6.145367412140575, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 7694 + }, + { + "epoch": 6.146166134185304, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 7695 + }, + { + "epoch": 6.146964856230032, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 7696 + }, + { + "epoch": 6.147763578274761, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 7697 + }, + { + "epoch": 6.1485623003194885, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 7698 + }, + { + "epoch": 6.149361022364217, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 7699 + }, + { + "epoch": 6.1501597444089455, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 7700 + }, + { + "epoch": 6.150958466453674, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 7701 + }, + { + "epoch": 6.151757188498403, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 7702 + }, + { + "epoch": 6.152555910543131, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7703 + }, + { + "epoch": 6.15335463258786, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 7704 + }, + { + "epoch": 6.154153354632588, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 7705 + }, + { + "epoch": 6.154952076677317, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 7706 + }, + { + "epoch": 6.155750798722044, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 7707 + }, + { + "epoch": 6.156549520766773, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 7708 + }, + { + "epoch": 6.157348242811501, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 7709 + }, + { + "epoch": 6.15814696485623, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 7710 + }, + { + "epoch": 6.1589456869009584, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 7711 + }, + { + "epoch": 6.159744408945687, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 7712 + }, + { + "epoch": 6.1605431309904155, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 7713 + }, + { + "epoch": 6.161341853035144, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 7714 + }, + { + "epoch": 6.162140575079873, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 7715 + }, + { + "epoch": 6.1629392971246, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 7716 + }, + { + "epoch": 6.163738019169329, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 7717 + }, + { + "epoch": 6.164536741214057, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 7718 + }, + { + "epoch": 6.165335463258786, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 7719 + }, + { + "epoch": 6.166134185303514, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 7720 + }, + { + "epoch": 6.166932907348243, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 7721 + }, + { + "epoch": 6.167731629392971, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 7722 + }, + { + "epoch": 6.1685303514377, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 7723 + }, + { + "epoch": 6.169329073482428, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 7724 + }, + { + "epoch": 6.170127795527157, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 7725 + }, + { + "epoch": 6.170926517571885, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 7726 + }, + { + "epoch": 6.171725239616613, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 7727 + }, + { + "epoch": 6.172523961661342, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 7728 + }, + { + "epoch": 6.17332268370607, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 7729 + }, + { + "epoch": 6.174121405750799, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 7730 + }, + { + "epoch": 6.174920127795527, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 7731 + }, + { + "epoch": 6.175718849840256, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 7732 + }, + { + "epoch": 6.176517571884984, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 7733 + }, + { + "epoch": 6.177316293929713, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 7734 + }, + { + "epoch": 6.178115015974441, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 7735 + }, + { + "epoch": 6.178913738019169, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 7736 + }, + { + "epoch": 6.1797124600638975, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 7737 + }, + { + "epoch": 6.180511182108626, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 7738 + }, + { + "epoch": 6.181309904153355, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 7739 + }, + { + "epoch": 6.182108626198083, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 7740 + }, + { + "epoch": 6.182907348242812, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 7741 + }, + { + "epoch": 6.18370607028754, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 7742 + }, + { + "epoch": 6.184504792332269, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 7743 + }, + { + "epoch": 6.185303514376997, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 7744 + }, + { + "epoch": 6.186102236421725, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 7745 + }, + { + "epoch": 6.186900958466453, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 7746 + }, + { + "epoch": 6.187699680511182, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1212, + "step": 7747 + }, + { + "epoch": 6.18849840255591, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 7748 + }, + { + "epoch": 6.189297124600639, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 7749 + }, + { + "epoch": 6.1900958466453675, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 7750 + }, + { + "epoch": 6.190894568690096, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 7751 + }, + { + "epoch": 6.1916932907348246, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 7752 + }, + { + "epoch": 6.192492012779553, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 7753 + }, + { + "epoch": 6.193290734824281, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 7754 + }, + { + "epoch": 6.194089456869009, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 7755 + }, + { + "epoch": 6.194888178913738, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 7756 + }, + { + "epoch": 6.195686900958466, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 7757 + }, + { + "epoch": 6.196485623003195, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 7758 + }, + { + "epoch": 6.197284345047923, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 7759 + }, + { + "epoch": 6.198083067092652, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 7760 + }, + { + "epoch": 6.19888178913738, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 7761 + }, + { + "epoch": 6.199680511182109, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 7762 + }, + { + "epoch": 6.2004792332268375, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 7763 + }, + { + "epoch": 6.201277955271565, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 7764 + }, + { + "epoch": 6.202076677316294, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 7765 + }, + { + "epoch": 6.202875399361022, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 7766 + }, + { + "epoch": 6.203674121405751, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 7767 + }, + { + "epoch": 6.204472843450479, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 7768 + }, + { + "epoch": 6.205271565495208, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 7769 + }, + { + "epoch": 6.206070287539936, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 7770 + }, + { + "epoch": 6.206869009584665, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 7771 + }, + { + "epoch": 6.207667731629393, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 7772 + }, + { + "epoch": 6.208466453674121, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7773 + }, + { + "epoch": 6.2092651757188495, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 7774 + }, + { + "epoch": 6.210063897763578, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7775 + }, + { + "epoch": 6.210862619808307, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 7776 + }, + { + "epoch": 6.211661341853035, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 7777 + }, + { + "epoch": 6.212460063897764, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 7778 + }, + { + "epoch": 6.213258785942492, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 7779 + }, + { + "epoch": 6.214057507987221, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 7780 + }, + { + "epoch": 6.214856230031949, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 7781 + }, + { + "epoch": 6.215654952076678, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 7782 + }, + { + "epoch": 6.216453674121405, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 7783 + }, + { + "epoch": 6.217252396166134, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 7784 + }, + { + "epoch": 6.218051118210862, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 7785 + }, + { + "epoch": 6.218849840255591, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 7786 + }, + { + "epoch": 6.2196485623003195, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 7787 + }, + { + "epoch": 6.220447284345048, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 7788 + }, + { + "epoch": 6.2212460063897765, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 7789 + }, + { + "epoch": 6.222044728434505, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 7790 + }, + { + "epoch": 6.222843450479234, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 7791 + }, + { + "epoch": 6.223642172523961, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 7792 + }, + { + "epoch": 6.22444089456869, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 7793 + }, + { + "epoch": 6.225239616613418, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 7794 + }, + { + "epoch": 6.226038338658147, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 7795 + }, + { + "epoch": 6.226837060702875, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 7796 + }, + { + "epoch": 6.227635782747604, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 7797 + }, + { + "epoch": 6.228434504792332, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 7798 + }, + { + "epoch": 6.229233226837061, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 7799 + }, + { + "epoch": 6.2300319488817895, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 7800 + }, + { + "epoch": 6.230830670926518, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 7801 + }, + { + "epoch": 6.231629392971246, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7802 + }, + { + "epoch": 6.232428115015974, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 7803 + }, + { + "epoch": 6.233226837060703, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 7804 + }, + { + "epoch": 6.234025559105431, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 7805 + }, + { + "epoch": 6.23482428115016, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 7806 + }, + { + "epoch": 6.235623003194888, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 7807 + }, + { + "epoch": 6.236421725239617, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 7808 + }, + { + "epoch": 6.237220447284345, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 7809 + }, + { + "epoch": 6.238019169329074, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 7810 + }, + { + "epoch": 6.2388178913738015, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 7811 + }, + { + "epoch": 6.23961661341853, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 7812 + }, + { + "epoch": 6.2404153354632586, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 7813 + }, + { + "epoch": 6.241214057507987, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 7814 + }, + { + "epoch": 6.242012779552716, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 7815 + }, + { + "epoch": 6.242811501597444, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7816 + }, + { + "epoch": 6.243610223642173, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 7817 + }, + { + "epoch": 6.244408945686901, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 7818 + }, + { + "epoch": 6.24520766773163, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 7819 + }, + { + "epoch": 6.246006389776358, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 7820 + }, + { + "epoch": 6.246805111821086, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 7821 + }, + { + "epoch": 6.247603833865814, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 7822 + }, + { + "epoch": 6.248402555910543, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 7823 + }, + { + "epoch": 6.2492012779552715, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 7824 + }, + { + "epoch": 6.25, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 7825 + }, + { + "epoch": 6.2507987220447285, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 7826 + }, + { + "epoch": 6.251597444089457, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 7827 + }, + { + "epoch": 6.252396166134186, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7828 + }, + { + "epoch": 6.253194888178914, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 7829 + }, + { + "epoch": 6.253993610223642, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 7830 + }, + { + "epoch": 6.25479233226837, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 7831 + }, + { + "epoch": 6.255591054313099, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 7832 + }, + { + "epoch": 6.256389776357827, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 7833 + }, + { + "epoch": 6.257188498402556, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 7834 + }, + { + "epoch": 6.257987220447284, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 7835 + }, + { + "epoch": 6.258785942492013, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 7836 + }, + { + "epoch": 6.2595846645367414, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 7837 + }, + { + "epoch": 6.26038338658147, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 7838 + }, + { + "epoch": 6.261182108626198, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 7839 + }, + { + "epoch": 6.261980830670926, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 7840 + }, + { + "epoch": 6.262779552715655, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 7841 + }, + { + "epoch": 6.263578274760383, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 7842 + }, + { + "epoch": 6.264376996805112, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7843 + }, + { + "epoch": 6.26517571884984, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 7844 + }, + { + "epoch": 6.265974440894569, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 7845 + }, + { + "epoch": 6.266773162939297, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1182, + "step": 7846 + }, + { + "epoch": 6.267571884984026, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 7847 + }, + { + "epoch": 6.268370607028754, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 7848 + }, + { + "epoch": 6.269169329073483, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 7849 + }, + { + "epoch": 6.2699680511182105, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 7850 + }, + { + "epoch": 6.270766773162939, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 7851 + }, + { + "epoch": 6.271565495207668, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 7852 + }, + { + "epoch": 6.272364217252396, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 7853 + }, + { + "epoch": 6.273162939297125, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 7854 + }, + { + "epoch": 6.273961661341853, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 7855 + }, + { + "epoch": 6.274760383386582, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 7856 + }, + { + "epoch": 6.27555910543131, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 7857 + }, + { + "epoch": 6.276357827476039, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 7858 + }, + { + "epoch": 6.277156549520766, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 7859 + }, + { + "epoch": 6.277955271565495, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 7860 + }, + { + "epoch": 6.2787539936102235, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 7861 + }, + { + "epoch": 6.279552715654952, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 7862 + }, + { + "epoch": 6.2803514376996805, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 7863 + }, + { + "epoch": 6.281150159744409, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 7864 + }, + { + "epoch": 6.281948881789138, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 7865 + }, + { + "epoch": 6.282747603833866, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 7866 + }, + { + "epoch": 6.283546325878595, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 7867 + }, + { + "epoch": 6.284345047923322, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 7868 + }, + { + "epoch": 6.285143769968051, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 7869 + }, + { + "epoch": 6.285942492012779, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 7870 + }, + { + "epoch": 6.286741214057508, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 7871 + }, + { + "epoch": 6.287539936102236, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 7872 + }, + { + "epoch": 6.288338658146965, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 7873 + }, + { + "epoch": 6.289137380191693, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 7874 + }, + { + "epoch": 6.289936102236422, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 7875 + }, + { + "epoch": 6.2907348242811505, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 7876 + }, + { + "epoch": 6.291533546325878, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 7877 + }, + { + "epoch": 6.292332268370607, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 7878 + }, + { + "epoch": 6.293130990415335, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 7879 + }, + { + "epoch": 6.293929712460064, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 7880 + }, + { + "epoch": 6.294728434504792, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 7881 + }, + { + "epoch": 6.295527156549521, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 7882 + }, + { + "epoch": 6.296325878594249, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 7883 + }, + { + "epoch": 6.297124600638978, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 7884 + }, + { + "epoch": 6.297923322683706, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 7885 + }, + { + "epoch": 6.298722044728435, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 7886 + }, + { + "epoch": 6.2995207667731625, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1223, + "step": 7887 + }, + { + "epoch": 6.300319488817891, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 7888 + }, + { + "epoch": 6.30111821086262, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 7889 + }, + { + "epoch": 6.301916932907348, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 7890 + }, + { + "epoch": 6.302715654952077, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 7891 + }, + { + "epoch": 6.303514376996805, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 7892 + }, + { + "epoch": 6.304313099041534, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1171, + "step": 7893 + }, + { + "epoch": 6.305111821086262, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 7894 + }, + { + "epoch": 6.305910543130991, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 7895 + }, + { + "epoch": 6.306709265175719, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 7896 + }, + { + "epoch": 6.307507987220447, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 7897 + }, + { + "epoch": 6.3083067092651754, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 7898 + }, + { + "epoch": 6.309105431309904, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 7899 + }, + { + "epoch": 6.3099041533546325, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 7900 + }, + { + "epoch": 6.310702875399361, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 7901 + }, + { + "epoch": 6.31150159744409, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 7902 + }, + { + "epoch": 6.312300319488818, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 7903 + }, + { + "epoch": 6.313099041533547, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 7904 + }, + { + "epoch": 6.313897763578275, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 7905 + }, + { + "epoch": 6.314696485623003, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 7906 + }, + { + "epoch": 6.315495207667731, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 7907 + }, + { + "epoch": 6.31629392971246, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7908 + }, + { + "epoch": 6.317092651757188, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 7909 + }, + { + "epoch": 6.317891373801917, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 7910 + }, + { + "epoch": 6.318690095846645, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 7911 + }, + { + "epoch": 6.319488817891374, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 7912 + }, + { + "epoch": 6.3202875399361025, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 7913 + }, + { + "epoch": 6.321086261980831, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 7914 + }, + { + "epoch": 6.321884984025559, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 7915 + }, + { + "epoch": 6.322683706070287, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 7916 + }, + { + "epoch": 6.323482428115016, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 7917 + }, + { + "epoch": 6.324281150159744, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 7918 + }, + { + "epoch": 6.325079872204473, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 7919 + }, + { + "epoch": 6.325878594249201, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 7920 + }, + { + "epoch": 6.32667731629393, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 7921 + }, + { + "epoch": 6.327476038338658, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 7922 + }, + { + "epoch": 6.328274760383387, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 7923 + }, + { + "epoch": 6.329073482428115, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 7924 + }, + { + "epoch": 6.329872204472843, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 7925 + }, + { + "epoch": 6.330670926517572, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 7926 + }, + { + "epoch": 6.3314696485623, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 7927 + }, + { + "epoch": 6.332268370607029, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 7928 + }, + { + "epoch": 6.333067092651757, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1166, + "step": 7929 + }, + { + "epoch": 6.333865814696486, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 7930 + }, + { + "epoch": 6.334664536741214, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1179, + "step": 7931 + }, + { + "epoch": 6.335463258785943, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 7932 + }, + { + "epoch": 6.336261980830671, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 7933 + }, + { + "epoch": 6.3370607028754, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 7934 + }, + { + "epoch": 6.337859424920127, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 7935 + }, + { + "epoch": 6.338658146964856, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 7936 + }, + { + "epoch": 6.3394568690095845, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 7937 + }, + { + "epoch": 6.340255591054313, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 7938 + }, + { + "epoch": 6.3410543130990416, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 7939 + }, + { + "epoch": 6.34185303514377, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 7940 + }, + { + "epoch": 6.342651757188499, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 7941 + }, + { + "epoch": 6.343450479233227, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 7942 + }, + { + "epoch": 6.344249201277956, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 7943 + }, + { + "epoch": 6.345047923322683, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 7944 + }, + { + "epoch": 6.345846645367412, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 7945 + }, + { + "epoch": 6.34664536741214, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 7946 + }, + { + "epoch": 6.347444089456869, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 7947 + }, + { + "epoch": 6.348242811501597, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 7948 + }, + { + "epoch": 6.349041533546326, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 7949 + }, + { + "epoch": 6.3498402555910545, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 7950 + }, + { + "epoch": 6.350638977635783, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 7951 + }, + { + "epoch": 6.3514376996805115, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 7952 + }, + { + "epoch": 6.352236421725239, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 7953 + }, + { + "epoch": 6.353035143769968, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 7954 + }, + { + "epoch": 6.353833865814696, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 7955 + }, + { + "epoch": 6.354632587859425, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7956 + }, + { + "epoch": 6.355431309904153, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 7957 + }, + { + "epoch": 6.356230031948882, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 7958 + }, + { + "epoch": 6.35702875399361, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 7959 + }, + { + "epoch": 6.357827476038339, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 7960 + }, + { + "epoch": 6.358626198083067, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 7961 + }, + { + "epoch": 6.359424920127796, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 7962 + }, + { + "epoch": 6.360223642172524, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 7963 + }, + { + "epoch": 6.361022364217252, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 7964 + }, + { + "epoch": 6.361821086261981, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 7965 + }, + { + "epoch": 6.362619808306709, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 7966 + }, + { + "epoch": 6.363418530351438, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 7967 + }, + { + "epoch": 6.364217252396166, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 7968 + }, + { + "epoch": 6.365015974440895, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 7969 + }, + { + "epoch": 6.365814696485623, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 7970 + }, + { + "epoch": 6.366613418530352, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 7971 + }, + { + "epoch": 6.36741214057508, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 7972 + }, + { + "epoch": 6.368210862619808, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 7973 + }, + { + "epoch": 6.3690095846645365, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 7974 + }, + { + "epoch": 6.369808306709265, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 7975 + }, + { + "epoch": 6.3706070287539935, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 7976 + }, + { + "epoch": 6.371405750798722, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 7977 + }, + { + "epoch": 6.372204472843451, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 7978 + }, + { + "epoch": 6.373003194888179, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 7979 + }, + { + "epoch": 6.373801916932908, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 7980 + }, + { + "epoch": 6.374600638977636, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 7981 + }, + { + "epoch": 6.375399361022364, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 7982 + }, + { + "epoch": 6.376198083067092, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 7983 + }, + { + "epoch": 6.376996805111821, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 7984 + }, + { + "epoch": 6.377795527156549, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 7985 + }, + { + "epoch": 6.378594249201278, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 7986 + }, + { + "epoch": 6.3793929712460065, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 7987 + }, + { + "epoch": 6.380191693290735, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 7988 + }, + { + "epoch": 6.3809904153354635, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 7989 + }, + { + "epoch": 6.381789137380192, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 7990 + }, + { + "epoch": 6.38258785942492, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 7991 + }, + { + "epoch": 6.383386581469648, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 7992 + }, + { + "epoch": 6.384185303514377, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 7993 + }, + { + "epoch": 6.384984025559105, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 7994 + }, + { + "epoch": 6.385782747603834, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 7995 + }, + { + "epoch": 6.386581469648562, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 7996 + }, + { + "epoch": 6.387380191693291, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 7997 + }, + { + "epoch": 6.388178913738019, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 7998 + }, + { + "epoch": 6.388977635782748, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 7999 + }, + { + "epoch": 6.389776357827476, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 8000 + }, + { + "epoch": 6.390575079872204, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 8001 + }, + { + "epoch": 6.391373801916933, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 8002 + }, + { + "epoch": 6.392172523961661, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 8003 + }, + { + "epoch": 6.39297124600639, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 8004 + }, + { + "epoch": 6.393769968051118, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 8005 + }, + { + "epoch": 6.394568690095847, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 8006 + }, + { + "epoch": 6.395367412140575, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 8007 + }, + { + "epoch": 6.396166134185304, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 8008 + }, + { + "epoch": 6.396964856230032, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 8009 + }, + { + "epoch": 6.397763578274761, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 8010 + }, + { + "epoch": 6.3985623003194885, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 8011 + }, + { + "epoch": 6.399361022364217, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 8012 + }, + { + "epoch": 6.4001597444089455, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 8013 + }, + { + "epoch": 6.400958466453674, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 8014 + }, + { + "epoch": 6.401757188498403, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 8015 + }, + { + "epoch": 6.402555910543131, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 8016 + }, + { + "epoch": 6.40335463258786, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8017 + }, + { + "epoch": 6.404153354632588, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 8018 + }, + { + "epoch": 6.404952076677317, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 8019 + }, + { + "epoch": 6.405750798722044, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 8020 + }, + { + "epoch": 6.406549520766773, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 8021 + }, + { + "epoch": 6.407348242811501, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 8022 + }, + { + "epoch": 6.40814696485623, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 8023 + }, + { + "epoch": 6.4089456869009584, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 8024 + }, + { + "epoch": 6.409744408945687, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 8025 + }, + { + "epoch": 6.4105431309904155, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 8026 + }, + { + "epoch": 6.411341853035144, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 8027 + }, + { + "epoch": 6.412140575079873, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 8028 + }, + { + "epoch": 6.4129392971246, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 8029 + }, + { + "epoch": 6.413738019169329, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 8030 + }, + { + "epoch": 6.414536741214057, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 8031 + }, + { + "epoch": 6.415335463258786, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 8032 + }, + { + "epoch": 6.416134185303514, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 8033 + }, + { + "epoch": 6.416932907348243, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 8034 + }, + { + "epoch": 6.417731629392971, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 8035 + }, + { + "epoch": 6.4185303514377, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 8036 + }, + { + "epoch": 6.419329073482428, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 8037 + }, + { + "epoch": 6.420127795527157, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 8038 + }, + { + "epoch": 6.420926517571885, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 8039 + }, + { + "epoch": 6.421725239616613, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 8040 + }, + { + "epoch": 6.422523961661342, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 8041 + }, + { + "epoch": 6.42332268370607, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 8042 + }, + { + "epoch": 6.424121405750799, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 8043 + }, + { + "epoch": 6.424920127795527, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 8044 + }, + { + "epoch": 6.425718849840256, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 8045 + }, + { + "epoch": 6.426517571884984, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 8046 + }, + { + "epoch": 6.427316293929713, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 8047 + }, + { + "epoch": 6.428115015974441, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 8048 + }, + { + "epoch": 6.428913738019169, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 8049 + }, + { + "epoch": 6.4297124600638975, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 8050 + }, + { + "epoch": 6.430511182108626, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 8051 + }, + { + "epoch": 6.431309904153355, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 8052 + }, + { + "epoch": 6.432108626198083, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 8053 + }, + { + "epoch": 6.432907348242812, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 8054 + }, + { + "epoch": 6.43370607028754, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 8055 + }, + { + "epoch": 6.434504792332269, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 8056 + }, + { + "epoch": 6.435303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 8057 + }, + { + "epoch": 6.436102236421725, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 8058 + }, + { + "epoch": 6.436900958466453, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 8059 + }, + { + "epoch": 6.437699680511182, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 8060 + }, + { + "epoch": 6.43849840255591, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 8061 + }, + { + "epoch": 6.439297124600639, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 8062 + }, + { + "epoch": 6.4400958466453675, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 8063 + }, + { + "epoch": 6.440894568690096, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 8064 + }, + { + "epoch": 6.4416932907348246, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 8065 + }, + { + "epoch": 6.442492012779553, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 8066 + }, + { + "epoch": 6.443290734824281, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 8067 + }, + { + "epoch": 6.444089456869009, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 8068 + }, + { + "epoch": 6.444888178913738, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 8069 + }, + { + "epoch": 6.445686900958466, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 8070 + }, + { + "epoch": 6.446485623003195, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 8071 + }, + { + "epoch": 6.447284345047923, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 8072 + }, + { + "epoch": 6.448083067092652, + "grad_norm": 0.484375, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 8073 + }, + { + "epoch": 6.44888178913738, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 8074 + }, + { + "epoch": 6.449680511182109, + "grad_norm": 0.318359375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 8075 + }, + { + "epoch": 6.4504792332268375, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 8076 + }, + { + "epoch": 6.451277955271565, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 8077 + }, + { + "epoch": 6.452076677316294, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 8078 + }, + { + "epoch": 6.452875399361022, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 8079 + }, + { + "epoch": 6.453674121405751, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 8080 + }, + { + "epoch": 6.454472843450479, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 8081 + }, + { + "epoch": 6.455271565495208, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8082 + }, + { + "epoch": 6.456070287539936, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 8083 + }, + { + "epoch": 6.456869009584665, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 8084 + }, + { + "epoch": 6.457667731629393, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 8085 + }, + { + "epoch": 6.458466453674122, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 8086 + }, + { + "epoch": 6.4592651757188495, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 8087 + }, + { + "epoch": 6.460063897763578, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 8088 + }, + { + "epoch": 6.460862619808307, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 8089 + }, + { + "epoch": 6.461661341853035, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 8090 + }, + { + "epoch": 6.462460063897764, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 8091 + }, + { + "epoch": 6.463258785942492, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 8092 + }, + { + "epoch": 6.464057507987221, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 8093 + }, + { + "epoch": 6.464856230031949, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 8094 + }, + { + "epoch": 6.465654952076678, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 8095 + }, + { + "epoch": 6.466453674121405, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 8096 + }, + { + "epoch": 6.467252396166134, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 8097 + }, + { + "epoch": 6.468051118210862, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 8098 + }, + { + "epoch": 6.468849840255591, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1185, + "step": 8099 + }, + { + "epoch": 6.4696485623003195, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 8100 + }, + { + "epoch": 6.470447284345048, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 8101 + }, + { + "epoch": 6.4712460063897765, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 8102 + }, + { + "epoch": 6.472044728434505, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 8103 + }, + { + "epoch": 6.472843450479234, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 8104 + }, + { + "epoch": 6.473642172523961, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 8105 + }, + { + "epoch": 6.47444089456869, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 8106 + }, + { + "epoch": 6.475239616613418, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 8107 + }, + { + "epoch": 6.476038338658147, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 8108 + }, + { + "epoch": 6.476837060702875, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 8109 + }, + { + "epoch": 6.477635782747604, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 8110 + }, + { + "epoch": 6.478434504792332, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 8111 + }, + { + "epoch": 6.479233226837061, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 8112 + }, + { + "epoch": 6.4800319488817895, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 8113 + }, + { + "epoch": 6.480830670926517, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 8114 + }, + { + "epoch": 6.481629392971246, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 8115 + }, + { + "epoch": 6.482428115015974, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8116 + }, + { + "epoch": 6.483226837060703, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 8117 + }, + { + "epoch": 6.484025559105431, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 8118 + }, + { + "epoch": 6.48482428115016, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 8119 + }, + { + "epoch": 6.485623003194888, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 8120 + }, + { + "epoch": 6.486421725239617, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 8121 + }, + { + "epoch": 6.487220447284345, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 8122 + }, + { + "epoch": 6.488019169329074, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 8123 + }, + { + "epoch": 6.488817891373802, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 8124 + }, + { + "epoch": 6.48961661341853, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 8125 + }, + { + "epoch": 6.4904153354632586, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 8126 + }, + { + "epoch": 6.491214057507987, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 8127 + }, + { + "epoch": 6.492012779552716, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 8128 + }, + { + "epoch": 6.492811501597444, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 8129 + }, + { + "epoch": 6.493610223642173, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 8130 + }, + { + "epoch": 6.494408945686901, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 8131 + }, + { + "epoch": 6.49520766773163, + "grad_norm": 0.09423828125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 8132 + }, + { + "epoch": 6.496006389776358, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 8133 + }, + { + "epoch": 6.496805111821086, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 8134 + }, + { + "epoch": 6.497603833865814, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 8135 + }, + { + "epoch": 6.498402555910543, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 8136 + }, + { + "epoch": 6.4992012779552715, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 8137 + }, + { + "epoch": 6.5, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 8138 + }, + { + "epoch": 6.5007987220447285, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 8139 + }, + { + "epoch": 6.501597444089457, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 8140 + }, + { + "epoch": 6.502396166134186, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 8141 + }, + { + "epoch": 6.503194888178914, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 8142 + }, + { + "epoch": 6.503993610223642, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 8143 + }, + { + "epoch": 6.50479233226837, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 8144 + }, + { + "epoch": 6.505591054313099, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 8145 + }, + { + "epoch": 6.506389776357827, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 8146 + }, + { + "epoch": 6.507188498402556, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 8147 + }, + { + "epoch": 6.507987220447284, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8148 + }, + { + "epoch": 6.508785942492013, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 8149 + }, + { + "epoch": 6.5095846645367414, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 8150 + }, + { + "epoch": 6.51038338658147, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 8151 + }, + { + "epoch": 6.511182108626198, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 8152 + }, + { + "epoch": 6.511980830670926, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 8153 + }, + { + "epoch": 6.512779552715655, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 8154 + }, + { + "epoch": 6.513578274760383, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 8155 + }, + { + "epoch": 6.514376996805112, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 8156 + }, + { + "epoch": 6.51517571884984, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 8157 + }, + { + "epoch": 6.515974440894569, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 8158 + }, + { + "epoch": 6.516773162939297, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 8159 + }, + { + "epoch": 6.517571884984026, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 8160 + }, + { + "epoch": 6.518370607028754, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 8161 + }, + { + "epoch": 6.519169329073483, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 8162 + }, + { + "epoch": 6.5199680511182105, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 8163 + }, + { + "epoch": 6.520766773162939, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 8164 + }, + { + "epoch": 6.521565495207668, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 8165 + }, + { + "epoch": 6.522364217252396, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 8166 + }, + { + "epoch": 6.523162939297125, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 8167 + }, + { + "epoch": 6.523961661341853, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 8168 + }, + { + "epoch": 6.524760383386582, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 8169 + }, + { + "epoch": 6.52555910543131, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 8170 + }, + { + "epoch": 6.526357827476039, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 8171 + }, + { + "epoch": 6.527156549520766, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 8172 + }, + { + "epoch": 6.527955271565495, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 8173 + }, + { + "epoch": 6.5287539936102235, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 8174 + }, + { + "epoch": 6.529552715654952, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 8175 + }, + { + "epoch": 6.5303514376996805, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 8176 + }, + { + "epoch": 6.531150159744409, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 8177 + }, + { + "epoch": 6.531948881789138, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.116, + "step": 8178 + }, + { + "epoch": 6.532747603833866, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 8179 + }, + { + "epoch": 6.533546325878595, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1204, + "step": 8180 + }, + { + "epoch": 6.534345047923322, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 8181 + }, + { + "epoch": 6.535143769968051, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 8182 + }, + { + "epoch": 6.535942492012779, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 8183 + }, + { + "epoch": 6.536741214057508, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 8184 + }, + { + "epoch": 6.537539936102236, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 8185 + }, + { + "epoch": 6.538338658146965, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 8186 + }, + { + "epoch": 6.539137380191693, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 8187 + }, + { + "epoch": 6.539936102236422, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 8188 + }, + { + "epoch": 6.5407348242811505, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 8189 + }, + { + "epoch": 6.541533546325878, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 8190 + }, + { + "epoch": 6.542332268370607, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 8191 + }, + { + "epoch": 6.543130990415335, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 8192 + }, + { + "epoch": 6.543929712460064, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 8193 + }, + { + "epoch": 6.544728434504792, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 8194 + }, + { + "epoch": 6.545527156549521, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 8195 + }, + { + "epoch": 6.546325878594249, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 8196 + }, + { + "epoch": 6.547124600638978, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 8197 + }, + { + "epoch": 6.547923322683706, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 8198 + }, + { + "epoch": 6.548722044728435, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 8199 + }, + { + "epoch": 6.549520766773163, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 8200 + }, + { + "epoch": 6.550319488817891, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 8201 + }, + { + "epoch": 6.55111821086262, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 8202 + }, + { + "epoch": 6.551916932907348, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 8203 + }, + { + "epoch": 6.552715654952077, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 8204 + }, + { + "epoch": 6.553514376996805, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 8205 + }, + { + "epoch": 6.554313099041534, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 8206 + }, + { + "epoch": 6.555111821086262, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 8207 + }, + { + "epoch": 6.555910543130991, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 8208 + }, + { + "epoch": 6.556709265175719, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 8209 + }, + { + "epoch": 6.557507987220447, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 8210 + }, + { + "epoch": 6.5583067092651754, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 8211 + }, + { + "epoch": 6.559105431309904, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 8212 + }, + { + "epoch": 6.5599041533546325, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 8213 + }, + { + "epoch": 6.560702875399361, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 8214 + }, + { + "epoch": 6.56150159744409, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8215 + }, + { + "epoch": 6.562300319488818, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 8216 + }, + { + "epoch": 6.563099041533547, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 8217 + }, + { + "epoch": 6.563897763578275, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 8218 + }, + { + "epoch": 6.564696485623003, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 8219 + }, + { + "epoch": 6.565495207667731, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 8220 + }, + { + "epoch": 6.56629392971246, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 8221 + }, + { + "epoch": 6.567092651757188, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 8222 + }, + { + "epoch": 6.567891373801917, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 8223 + }, + { + "epoch": 6.568690095846645, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 8224 + }, + { + "epoch": 6.569488817891374, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 8225 + }, + { + "epoch": 6.5702875399361025, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 8226 + }, + { + "epoch": 6.571086261980831, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 8227 + }, + { + "epoch": 6.571884984025559, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 8228 + }, + { + "epoch": 6.572683706070287, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 8229 + }, + { + "epoch": 6.573482428115016, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 8230 + }, + { + "epoch": 6.574281150159744, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 8231 + }, + { + "epoch": 6.575079872204473, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 8232 + }, + { + "epoch": 6.575878594249201, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 8233 + }, + { + "epoch": 6.57667731629393, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 8234 + }, + { + "epoch": 6.577476038338658, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 8235 + }, + { + "epoch": 6.578274760383387, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 8236 + }, + { + "epoch": 6.5790734824281145, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 8237 + }, + { + "epoch": 6.579872204472844, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 8238 + }, + { + "epoch": 6.580670926517572, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 8239 + }, + { + "epoch": 6.5814696485623, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8240 + }, + { + "epoch": 6.582268370607029, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 8241 + }, + { + "epoch": 6.583067092651757, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 8242 + }, + { + "epoch": 6.583865814696486, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 8243 + }, + { + "epoch": 6.584664536741214, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 8244 + }, + { + "epoch": 6.585463258785943, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 8245 + }, + { + "epoch": 6.586261980830671, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 8246 + }, + { + "epoch": 6.5870607028754, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 8247 + }, + { + "epoch": 6.587859424920127, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 8248 + }, + { + "epoch": 6.588658146964856, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 8249 + }, + { + "epoch": 6.5894568690095845, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 8250 + }, + { + "epoch": 6.590255591054313, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 8251 + }, + { + "epoch": 6.5910543130990416, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 8252 + }, + { + "epoch": 6.59185303514377, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 8253 + }, + { + "epoch": 6.592651757188499, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 8254 + }, + { + "epoch": 6.593450479233227, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 8255 + }, + { + "epoch": 6.594249201277956, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 8256 + }, + { + "epoch": 6.595047923322683, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 8257 + }, + { + "epoch": 6.595846645367412, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 8258 + }, + { + "epoch": 6.59664536741214, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 8259 + }, + { + "epoch": 6.597444089456869, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 8260 + }, + { + "epoch": 6.598242811501597, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 8261 + }, + { + "epoch": 6.599041533546326, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1213, + "step": 8262 + }, + { + "epoch": 6.5998402555910545, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 8263 + }, + { + "epoch": 6.600638977635783, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8264 + }, + { + "epoch": 6.6014376996805115, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 8265 + }, + { + "epoch": 6.602236421725239, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 8266 + }, + { + "epoch": 6.603035143769968, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 8267 + }, + { + "epoch": 6.603833865814696, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 8268 + }, + { + "epoch": 6.604632587859425, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8269 + }, + { + "epoch": 6.605431309904153, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 8270 + }, + { + "epoch": 6.606230031948882, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 8271 + }, + { + "epoch": 6.60702875399361, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 8272 + }, + { + "epoch": 6.607827476038339, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 8273 + }, + { + "epoch": 6.608626198083067, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 8274 + }, + { + "epoch": 6.609424920127795, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 8275 + }, + { + "epoch": 6.6102236421725244, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 8276 + }, + { + "epoch": 6.611022364217252, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 8277 + }, + { + "epoch": 6.611821086261981, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 8278 + }, + { + "epoch": 6.612619808306709, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 8279 + }, + { + "epoch": 6.613418530351438, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1221, + "step": 8280 + }, + { + "epoch": 6.614217252396166, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 8281 + }, + { + "epoch": 6.615015974440895, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 8282 + }, + { + "epoch": 6.615814696485623, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 8283 + }, + { + "epoch": 6.616613418530352, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 8284 + }, + { + "epoch": 6.61741214057508, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 8285 + }, + { + "epoch": 6.618210862619808, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 8286 + }, + { + "epoch": 6.6190095846645365, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 8287 + }, + { + "epoch": 6.619808306709265, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 8288 + }, + { + "epoch": 6.6206070287539935, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 8289 + }, + { + "epoch": 6.621405750798722, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 8290 + }, + { + "epoch": 6.622204472843451, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 8291 + }, + { + "epoch": 6.623003194888179, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 8292 + }, + { + "epoch": 6.623801916932908, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 8293 + }, + { + "epoch": 6.624600638977636, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 8294 + }, + { + "epoch": 6.625399361022364, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 8295 + }, + { + "epoch": 6.626198083067092, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 8296 + }, + { + "epoch": 6.626996805111821, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 8297 + }, + { + "epoch": 6.627795527156549, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 8298 + }, + { + "epoch": 6.628594249201278, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 8299 + }, + { + "epoch": 6.6293929712460065, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 8300 + }, + { + "epoch": 6.630191693290735, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 8301 + }, + { + "epoch": 6.6309904153354635, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 8302 + }, + { + "epoch": 6.631789137380192, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 8303 + }, + { + "epoch": 6.63258785942492, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 8304 + }, + { + "epoch": 6.633386581469648, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 8305 + }, + { + "epoch": 6.634185303514377, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 8306 + }, + { + "epoch": 6.634984025559105, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 8307 + }, + { + "epoch": 6.635782747603834, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 8308 + }, + { + "epoch": 6.636581469648562, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 8309 + }, + { + "epoch": 6.637380191693291, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 8310 + }, + { + "epoch": 6.638178913738019, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8311 + }, + { + "epoch": 6.638977635782748, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 8312 + }, + { + "epoch": 6.6397763578274756, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 8313 + }, + { + "epoch": 6.640575079872205, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 8314 + }, + { + "epoch": 6.641373801916933, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 8315 + }, + { + "epoch": 6.642172523961661, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 8316 + }, + { + "epoch": 6.64297124600639, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 8317 + }, + { + "epoch": 6.643769968051118, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 8318 + }, + { + "epoch": 6.644568690095847, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 8319 + }, + { + "epoch": 6.645367412140575, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 8320 + }, + { + "epoch": 6.646166134185304, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 8321 + }, + { + "epoch": 6.646964856230032, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 8322 + }, + { + "epoch": 6.647763578274761, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 8323 + }, + { + "epoch": 6.6485623003194885, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 8324 + }, + { + "epoch": 6.649361022364217, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 8325 + }, + { + "epoch": 6.6501597444089455, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 8326 + }, + { + "epoch": 6.650958466453674, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 8327 + }, + { + "epoch": 6.651757188498403, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 8328 + }, + { + "epoch": 6.652555910543131, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 8329 + }, + { + "epoch": 6.65335463258786, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 8330 + }, + { + "epoch": 6.654153354632588, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 8331 + }, + { + "epoch": 6.654952076677317, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 8332 + }, + { + "epoch": 6.655750798722044, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 8333 + }, + { + "epoch": 6.656549520766773, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 8334 + }, + { + "epoch": 6.657348242811501, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 8335 + }, + { + "epoch": 6.65814696485623, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 8336 + }, + { + "epoch": 6.6589456869009584, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 8337 + }, + { + "epoch": 6.659744408945687, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 8338 + }, + { + "epoch": 6.6605431309904155, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 8339 + }, + { + "epoch": 6.661341853035144, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 8340 + }, + { + "epoch": 6.662140575079873, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 8341 + }, + { + "epoch": 6.6629392971246, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 8342 + }, + { + "epoch": 6.663738019169329, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 8343 + }, + { + "epoch": 6.664536741214057, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 8344 + }, + { + "epoch": 6.665335463258786, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 8345 + }, + { + "epoch": 6.666134185303514, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 8346 + }, + { + "epoch": 6.666932907348243, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 8347 + }, + { + "epoch": 6.667731629392971, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 8348 + }, + { + "epoch": 6.6685303514377, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 8349 + }, + { + "epoch": 6.669329073482428, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 8350 + }, + { + "epoch": 6.670127795527156, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 8351 + }, + { + "epoch": 6.6709265175718855, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 8352 + }, + { + "epoch": 6.671725239616613, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 8353 + }, + { + "epoch": 6.672523961661342, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8354 + }, + { + "epoch": 6.67332268370607, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8355 + }, + { + "epoch": 6.674121405750799, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 8356 + }, + { + "epoch": 6.674920127795527, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 8357 + }, + { + "epoch": 6.675718849840256, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 8358 + }, + { + "epoch": 6.676517571884984, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 8359 + }, + { + "epoch": 6.677316293929713, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 8360 + }, + { + "epoch": 6.678115015974441, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 8361 + }, + { + "epoch": 6.678913738019169, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8362 + }, + { + "epoch": 6.6797124600638975, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 8363 + }, + { + "epoch": 6.680511182108626, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 8364 + }, + { + "epoch": 6.681309904153355, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 8365 + }, + { + "epoch": 6.682108626198083, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 8366 + }, + { + "epoch": 6.682907348242812, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 8367 + }, + { + "epoch": 6.68370607028754, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 8368 + }, + { + "epoch": 6.684504792332269, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 8369 + }, + { + "epoch": 6.685303514376997, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 8370 + }, + { + "epoch": 6.686102236421725, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 8371 + }, + { + "epoch": 6.686900958466453, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 8372 + }, + { + "epoch": 6.687699680511182, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 8373 + }, + { + "epoch": 6.68849840255591, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 8374 + }, + { + "epoch": 6.689297124600639, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 8375 + }, + { + "epoch": 6.6900958466453675, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 8376 + }, + { + "epoch": 6.690894568690096, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 8377 + }, + { + "epoch": 6.6916932907348246, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 8378 + }, + { + "epoch": 6.692492012779553, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 8379 + }, + { + "epoch": 6.693290734824281, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 8380 + }, + { + "epoch": 6.694089456869009, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 8381 + }, + { + "epoch": 6.694888178913738, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 8382 + }, + { + "epoch": 6.695686900958466, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 8383 + }, + { + "epoch": 6.696485623003195, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 8384 + }, + { + "epoch": 6.697284345047923, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 8385 + }, + { + "epoch": 6.698083067092652, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 8386 + }, + { + "epoch": 6.69888178913738, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 8387 + }, + { + "epoch": 6.699680511182109, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 8388 + }, + { + "epoch": 6.700479233226837, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 8389 + }, + { + "epoch": 6.701277955271565, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 8390 + }, + { + "epoch": 6.702076677316294, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 8391 + }, + { + "epoch": 6.702875399361022, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 8392 + }, + { + "epoch": 6.703674121405751, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 8393 + }, + { + "epoch": 6.704472843450479, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 8394 + }, + { + "epoch": 6.705271565495208, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 8395 + }, + { + "epoch": 6.706070287539936, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 8396 + }, + { + "epoch": 6.706869009584665, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 8397 + }, + { + "epoch": 6.707667731629393, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 8398 + }, + { + "epoch": 6.708466453674122, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 8399 + }, + { + "epoch": 6.7092651757188495, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 8400 + }, + { + "epoch": 6.710063897763578, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 8401 + }, + { + "epoch": 6.710862619808307, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 8402 + }, + { + "epoch": 6.711661341853035, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 8403 + }, + { + "epoch": 6.712460063897764, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 8404 + }, + { + "epoch": 6.713258785942492, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 8405 + }, + { + "epoch": 6.714057507987221, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8406 + }, + { + "epoch": 6.714856230031949, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 8407 + }, + { + "epoch": 6.715654952076678, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 8408 + }, + { + "epoch": 6.716453674121405, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 8409 + }, + { + "epoch": 6.717252396166134, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 8410 + }, + { + "epoch": 6.718051118210862, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 8411 + }, + { + "epoch": 6.718849840255591, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 8412 + }, + { + "epoch": 6.7196485623003195, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 8413 + }, + { + "epoch": 6.720447284345048, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 8414 + }, + { + "epoch": 6.7212460063897765, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 8415 + }, + { + "epoch": 6.722044728434505, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 8416 + }, + { + "epoch": 6.722843450479234, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 8417 + }, + { + "epoch": 6.723642172523961, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 8418 + }, + { + "epoch": 6.72444089456869, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 8419 + }, + { + "epoch": 6.725239616613418, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 8420 + }, + { + "epoch": 6.726038338658147, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 8421 + }, + { + "epoch": 6.726837060702875, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 8422 + }, + { + "epoch": 6.727635782747604, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 8423 + }, + { + "epoch": 6.728434504792332, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 8424 + }, + { + "epoch": 6.729233226837061, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 8425 + }, + { + "epoch": 6.7300319488817895, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 8426 + }, + { + "epoch": 6.730830670926517, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 8427 + }, + { + "epoch": 6.731629392971246, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 8428 + }, + { + "epoch": 6.732428115015974, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 8429 + }, + { + "epoch": 6.733226837060703, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 8430 + }, + { + "epoch": 6.734025559105431, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 8431 + }, + { + "epoch": 6.73482428115016, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 8432 + }, + { + "epoch": 6.735623003194888, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 8433 + }, + { + "epoch": 6.736421725239617, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 8434 + }, + { + "epoch": 6.737220447284345, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 8435 + }, + { + "epoch": 6.738019169329074, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 8436 + }, + { + "epoch": 6.738817891373802, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 8437 + }, + { + "epoch": 6.73961661341853, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 8438 + }, + { + "epoch": 6.7404153354632586, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 8439 + }, + { + "epoch": 6.741214057507987, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 8440 + }, + { + "epoch": 6.742012779552716, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 8441 + }, + { + "epoch": 6.742811501597444, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 8442 + }, + { + "epoch": 6.743610223642173, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 8443 + }, + { + "epoch": 6.744408945686901, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 8444 + }, + { + "epoch": 6.74520766773163, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 8445 + }, + { + "epoch": 6.746006389776358, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 8446 + }, + { + "epoch": 6.746805111821086, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 8447 + }, + { + "epoch": 6.747603833865814, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 8448 + }, + { + "epoch": 6.748402555910543, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 8449 + }, + { + "epoch": 6.7492012779552715, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 8450 + }, + { + "epoch": 6.75, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 8451 + }, + { + "epoch": 6.7507987220447285, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 8452 + }, + { + "epoch": 6.751597444089457, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 8453 + }, + { + "epoch": 6.752396166134186, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 8454 + }, + { + "epoch": 6.753194888178914, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 8455 + }, + { + "epoch": 6.753993610223642, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 8456 + }, + { + "epoch": 6.75479233226837, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 8457 + }, + { + "epoch": 6.755591054313099, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 8458 + }, + { + "epoch": 6.756389776357827, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 8459 + }, + { + "epoch": 6.757188498402556, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 8460 + }, + { + "epoch": 6.757987220447284, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 8461 + }, + { + "epoch": 6.758785942492013, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 8462 + }, + { + "epoch": 6.7595846645367414, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 8463 + }, + { + "epoch": 6.76038338658147, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 8464 + }, + { + "epoch": 6.761182108626198, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 8465 + }, + { + "epoch": 6.761980830670926, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 8466 + }, + { + "epoch": 6.762779552715655, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1244, + "step": 8467 + }, + { + "epoch": 6.763578274760383, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 8468 + }, + { + "epoch": 6.764376996805112, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 8469 + }, + { + "epoch": 6.76517571884984, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 8470 + }, + { + "epoch": 6.765974440894569, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 8471 + }, + { + "epoch": 6.766773162939297, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 8472 + }, + { + "epoch": 6.767571884984026, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 8473 + }, + { + "epoch": 6.768370607028754, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 8474 + }, + { + "epoch": 6.769169329073483, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 8475 + }, + { + "epoch": 6.7699680511182105, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 8476 + }, + { + "epoch": 6.770766773162939, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 8477 + }, + { + "epoch": 6.771565495207668, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 8478 + }, + { + "epoch": 6.772364217252396, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 8479 + }, + { + "epoch": 6.773162939297125, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 8480 + }, + { + "epoch": 6.773961661341853, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 8481 + }, + { + "epoch": 6.774760383386582, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 8482 + }, + { + "epoch": 6.77555910543131, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 8483 + }, + { + "epoch": 6.776357827476039, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 8484 + }, + { + "epoch": 6.777156549520766, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 8485 + }, + { + "epoch": 6.777955271565495, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 8486 + }, + { + "epoch": 6.7787539936102235, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 8487 + }, + { + "epoch": 6.779552715654952, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 8488 + }, + { + "epoch": 6.7803514376996805, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 8489 + }, + { + "epoch": 6.781150159744409, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 8490 + }, + { + "epoch": 6.781948881789138, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 8491 + }, + { + "epoch": 6.782747603833866, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 8492 + }, + { + "epoch": 6.783546325878595, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 8493 + }, + { + "epoch": 6.784345047923322, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 8494 + }, + { + "epoch": 6.785143769968051, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 8495 + }, + { + "epoch": 6.785942492012779, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 8496 + }, + { + "epoch": 6.786741214057508, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 8497 + }, + { + "epoch": 6.787539936102236, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8498 + }, + { + "epoch": 6.788338658146965, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 8499 + }, + { + "epoch": 6.789137380191693, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 8500 + }, + { + "epoch": 6.789936102236422, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 8501 + }, + { + "epoch": 6.7907348242811505, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 8502 + }, + { + "epoch": 6.791533546325878, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 8503 + }, + { + "epoch": 6.792332268370607, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 8504 + }, + { + "epoch": 6.793130990415335, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 8505 + }, + { + "epoch": 6.793929712460064, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 8506 + }, + { + "epoch": 6.794728434504792, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 8507 + }, + { + "epoch": 6.795527156549521, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 8508 + }, + { + "epoch": 6.796325878594249, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 8509 + }, + { + "epoch": 6.797124600638978, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 8510 + }, + { + "epoch": 6.797923322683706, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 8511 + }, + { + "epoch": 6.798722044728435, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 8512 + }, + { + "epoch": 6.799520766773163, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 8513 + }, + { + "epoch": 6.800319488817891, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1177, + "step": 8514 + }, + { + "epoch": 6.80111821086262, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8515 + }, + { + "epoch": 6.801916932907348, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 8516 + }, + { + "epoch": 6.802715654952077, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 8517 + }, + { + "epoch": 6.803514376996805, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8518 + }, + { + "epoch": 6.804313099041534, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 8519 + }, + { + "epoch": 6.805111821086262, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 8520 + }, + { + "epoch": 6.805910543130991, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 8521 + }, + { + "epoch": 6.806709265175719, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 8522 + }, + { + "epoch": 6.807507987220447, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 8523 + }, + { + "epoch": 6.8083067092651754, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 8524 + }, + { + "epoch": 6.809105431309904, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 8525 + }, + { + "epoch": 6.8099041533546325, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 8526 + }, + { + "epoch": 6.810702875399361, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8527 + }, + { + "epoch": 6.81150159744409, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 8528 + }, + { + "epoch": 6.812300319488818, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1202, + "step": 8529 + }, + { + "epoch": 6.813099041533547, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8530 + }, + { + "epoch": 6.813897763578275, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8531 + }, + { + "epoch": 6.814696485623003, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 8532 + }, + { + "epoch": 6.815495207667731, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 8533 + }, + { + "epoch": 6.81629392971246, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 8534 + }, + { + "epoch": 6.817092651757188, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 8535 + }, + { + "epoch": 6.817891373801917, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 8536 + }, + { + "epoch": 6.818690095846645, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 8537 + }, + { + "epoch": 6.819488817891374, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 8538 + }, + { + "epoch": 6.8202875399361025, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 8539 + }, + { + "epoch": 6.821086261980831, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 8540 + }, + { + "epoch": 6.821884984025559, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 8541 + }, + { + "epoch": 6.822683706070287, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 8542 + }, + { + "epoch": 6.823482428115016, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 8543 + }, + { + "epoch": 6.824281150159744, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 8544 + }, + { + "epoch": 6.825079872204473, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 8545 + }, + { + "epoch": 6.825878594249201, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8546 + }, + { + "epoch": 6.82667731629393, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 8547 + }, + { + "epoch": 6.827476038338658, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 8548 + }, + { + "epoch": 6.828274760383387, + "grad_norm": 0.337890625, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 8549 + }, + { + "epoch": 6.8290734824281145, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 8550 + }, + { + "epoch": 6.829872204472844, + "grad_norm": 0.34765625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 8551 + }, + { + "epoch": 6.830670926517572, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 8552 + }, + { + "epoch": 6.8314696485623, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 8553 + }, + { + "epoch": 6.832268370607029, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 8554 + }, + { + "epoch": 6.833067092651757, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 8555 + }, + { + "epoch": 6.833865814696486, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 8556 + }, + { + "epoch": 6.834664536741214, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 8557 + }, + { + "epoch": 6.835463258785943, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 8558 + }, + { + "epoch": 6.836261980830671, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 8559 + }, + { + "epoch": 6.8370607028754, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 8560 + }, + { + "epoch": 6.837859424920127, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 8561 + }, + { + "epoch": 6.838658146964856, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 8562 + }, + { + "epoch": 6.8394568690095845, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 8563 + }, + { + "epoch": 6.840255591054313, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 8564 + }, + { + "epoch": 6.8410543130990416, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 8565 + }, + { + "epoch": 6.84185303514377, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 8566 + }, + { + "epoch": 6.842651757188499, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 8567 + }, + { + "epoch": 6.843450479233227, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 8568 + }, + { + "epoch": 6.844249201277956, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 8569 + }, + { + "epoch": 6.845047923322683, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 8570 + }, + { + "epoch": 6.845846645367412, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 8571 + }, + { + "epoch": 6.84664536741214, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 8572 + }, + { + "epoch": 6.847444089456869, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 8573 + }, + { + "epoch": 6.848242811501597, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 8574 + }, + { + "epoch": 6.849041533546326, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 8575 + }, + { + "epoch": 6.8498402555910545, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 8576 + }, + { + "epoch": 6.850638977635783, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 8577 + }, + { + "epoch": 6.8514376996805115, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 8578 + }, + { + "epoch": 6.852236421725239, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 8579 + }, + { + "epoch": 6.853035143769968, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 8580 + }, + { + "epoch": 6.853833865814696, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 8581 + }, + { + "epoch": 6.854632587859425, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 8582 + }, + { + "epoch": 6.855431309904153, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 8583 + }, + { + "epoch": 6.856230031948882, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 8584 + }, + { + "epoch": 6.85702875399361, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 8585 + }, + { + "epoch": 6.857827476038339, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 8586 + }, + { + "epoch": 6.858626198083067, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 8587 + }, + { + "epoch": 6.859424920127795, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 8588 + }, + { + "epoch": 6.8602236421725244, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 8589 + }, + { + "epoch": 6.861022364217252, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 8590 + }, + { + "epoch": 6.861821086261981, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 8591 + }, + { + "epoch": 6.862619808306709, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 8592 + }, + { + "epoch": 6.863418530351438, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 8593 + }, + { + "epoch": 6.864217252396166, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 8594 + }, + { + "epoch": 6.865015974440895, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 8595 + }, + { + "epoch": 6.865814696485623, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 8596 + }, + { + "epoch": 6.866613418530352, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 8597 + }, + { + "epoch": 6.86741214057508, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 8598 + }, + { + "epoch": 6.868210862619808, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 8599 + }, + { + "epoch": 6.8690095846645365, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 8600 + }, + { + "epoch": 6.869808306709265, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 8601 + }, + { + "epoch": 6.8706070287539935, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 8602 + }, + { + "epoch": 6.871405750798722, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 8603 + }, + { + "epoch": 6.872204472843451, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.119, + "step": 8604 + }, + { + "epoch": 6.873003194888179, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 8605 + }, + { + "epoch": 6.873801916932908, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 8606 + }, + { + "epoch": 6.874600638977636, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8607 + }, + { + "epoch": 6.875399361022364, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 8608 + }, + { + "epoch": 6.876198083067092, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 8609 + }, + { + "epoch": 6.876996805111821, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 8610 + }, + { + "epoch": 6.877795527156549, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 8611 + }, + { + "epoch": 6.878594249201278, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 8612 + }, + { + "epoch": 6.8793929712460065, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 8613 + }, + { + "epoch": 6.880191693290735, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 8614 + }, + { + "epoch": 6.8809904153354635, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 8615 + }, + { + "epoch": 6.881789137380192, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 8616 + }, + { + "epoch": 6.88258785942492, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 8617 + }, + { + "epoch": 6.883386581469648, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 8618 + }, + { + "epoch": 6.884185303514377, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 8619 + }, + { + "epoch": 6.884984025559105, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 8620 + }, + { + "epoch": 6.885782747603834, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 8621 + }, + { + "epoch": 6.886581469648562, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 8622 + }, + { + "epoch": 6.887380191693291, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 8623 + }, + { + "epoch": 6.888178913738019, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 8624 + }, + { + "epoch": 6.888977635782748, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 8625 + }, + { + "epoch": 6.8897763578274756, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 8626 + }, + { + "epoch": 6.890575079872205, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 8627 + }, + { + "epoch": 6.891373801916933, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 8628 + }, + { + "epoch": 6.892172523961661, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 8629 + }, + { + "epoch": 6.89297124600639, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 8630 + }, + { + "epoch": 6.893769968051118, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 8631 + }, + { + "epoch": 6.894568690095847, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 8632 + }, + { + "epoch": 6.895367412140575, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 8633 + }, + { + "epoch": 6.896166134185304, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 8634 + }, + { + "epoch": 6.896964856230032, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 8635 + }, + { + "epoch": 6.897763578274761, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 8636 + }, + { + "epoch": 6.8985623003194885, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8637 + }, + { + "epoch": 6.899361022364217, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 8638 + }, + { + "epoch": 6.9001597444089455, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 8639 + }, + { + "epoch": 6.900958466453674, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 8640 + }, + { + "epoch": 6.901757188498403, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 8641 + }, + { + "epoch": 6.902555910543131, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 8642 + }, + { + "epoch": 6.90335463258786, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 8643 + }, + { + "epoch": 6.904153354632588, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 8644 + }, + { + "epoch": 6.904952076677317, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 8645 + }, + { + "epoch": 6.905750798722044, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 8646 + }, + { + "epoch": 6.906549520766773, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 8647 + }, + { + "epoch": 6.907348242811501, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 8648 + }, + { + "epoch": 6.90814696485623, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 8649 + }, + { + "epoch": 6.9089456869009584, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 8650 + }, + { + "epoch": 6.909744408945687, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 8651 + }, + { + "epoch": 6.9105431309904155, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 8652 + }, + { + "epoch": 6.911341853035144, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 8653 + }, + { + "epoch": 6.912140575079873, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 8654 + }, + { + "epoch": 6.9129392971246, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 8655 + }, + { + "epoch": 6.913738019169329, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 8656 + }, + { + "epoch": 6.914536741214057, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 8657 + }, + { + "epoch": 6.915335463258786, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 8658 + }, + { + "epoch": 6.916134185303514, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8659 + }, + { + "epoch": 6.916932907348243, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 8660 + }, + { + "epoch": 6.917731629392971, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 8661 + }, + { + "epoch": 6.9185303514377, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 8662 + }, + { + "epoch": 6.919329073482428, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 8663 + }, + { + "epoch": 6.920127795527156, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 8664 + }, + { + "epoch": 6.9209265175718855, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 8665 + }, + { + "epoch": 6.921725239616613, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 8666 + }, + { + "epoch": 6.922523961661342, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 8667 + }, + { + "epoch": 6.92332268370607, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 8668 + }, + { + "epoch": 6.924121405750799, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 8669 + }, + { + "epoch": 6.924920127795527, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 8670 + }, + { + "epoch": 6.925718849840256, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 8671 + }, + { + "epoch": 6.926517571884984, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 8672 + }, + { + "epoch": 6.927316293929713, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 8673 + }, + { + "epoch": 6.928115015974441, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 8674 + }, + { + "epoch": 6.928913738019169, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 8675 + }, + { + "epoch": 6.9297124600638975, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 8676 + }, + { + "epoch": 6.930511182108626, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 8677 + }, + { + "epoch": 6.931309904153355, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 8678 + }, + { + "epoch": 6.932108626198083, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 8679 + }, + { + "epoch": 6.932907348242812, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 8680 + }, + { + "epoch": 6.93370607028754, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 8681 + }, + { + "epoch": 6.934504792332269, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 8682 + }, + { + "epoch": 6.935303514376997, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 8683 + }, + { + "epoch": 6.936102236421725, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 8684 + }, + { + "epoch": 6.936900958466453, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 8685 + }, + { + "epoch": 6.937699680511182, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 8686 + }, + { + "epoch": 6.93849840255591, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 8687 + }, + { + "epoch": 6.939297124600639, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 8688 + }, + { + "epoch": 6.9400958466453675, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 8689 + }, + { + "epoch": 6.940894568690096, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8690 + }, + { + "epoch": 6.9416932907348246, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 8691 + }, + { + "epoch": 6.942492012779553, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 8692 + }, + { + "epoch": 6.943290734824281, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 8693 + }, + { + "epoch": 6.944089456869009, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 8694 + }, + { + "epoch": 6.944888178913738, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 8695 + }, + { + "epoch": 6.945686900958466, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 8696 + }, + { + "epoch": 6.946485623003195, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 8697 + }, + { + "epoch": 6.947284345047923, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 8698 + }, + { + "epoch": 6.948083067092652, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 8699 + }, + { + "epoch": 6.94888178913738, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 8700 + }, + { + "epoch": 6.949680511182109, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 8701 + }, + { + "epoch": 6.950479233226837, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 8702 + }, + { + "epoch": 6.951277955271565, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 8703 + }, + { + "epoch": 6.952076677316294, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 8704 + }, + { + "epoch": 6.952875399361022, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 8705 + }, + { + "epoch": 6.953674121405751, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 8706 + }, + { + "epoch": 6.954472843450479, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 8707 + }, + { + "epoch": 6.955271565495208, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 8708 + }, + { + "epoch": 6.956070287539936, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 8709 + }, + { + "epoch": 6.956869009584665, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 8710 + }, + { + "epoch": 6.957667731629393, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8711 + }, + { + "epoch": 6.958466453674122, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8712 + }, + { + "epoch": 6.9592651757188495, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 8713 + }, + { + "epoch": 6.960063897763578, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 8714 + }, + { + "epoch": 6.960862619808307, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 8715 + }, + { + "epoch": 6.961661341853035, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 8716 + }, + { + "epoch": 6.962460063897764, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 8717 + }, + { + "epoch": 6.963258785942492, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 8718 + }, + { + "epoch": 6.964057507987221, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 8719 + }, + { + "epoch": 6.964856230031949, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 8720 + }, + { + "epoch": 6.965654952076678, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 8721 + }, + { + "epoch": 6.966453674121405, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 8722 + }, + { + "epoch": 6.967252396166134, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 8723 + }, + { + "epoch": 6.968051118210862, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 8724 + }, + { + "epoch": 6.968849840255591, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 8725 + }, + { + "epoch": 6.9696485623003195, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 8726 + }, + { + "epoch": 6.970447284345048, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 8727 + }, + { + "epoch": 6.9712460063897765, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 8728 + }, + { + "epoch": 6.972044728434505, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 8729 + }, + { + "epoch": 6.972843450479234, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 8730 + }, + { + "epoch": 6.973642172523961, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 8731 + }, + { + "epoch": 6.97444089456869, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 8732 + }, + { + "epoch": 6.975239616613418, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 8733 + }, + { + "epoch": 6.976038338658147, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 8734 + }, + { + "epoch": 6.976837060702875, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 8735 + }, + { + "epoch": 6.977635782747604, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 8736 + }, + { + "epoch": 6.978434504792332, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 8737 + }, + { + "epoch": 6.979233226837061, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 8738 + }, + { + "epoch": 6.9800319488817895, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 8739 + }, + { + "epoch": 6.980830670926517, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 8740 + }, + { + "epoch": 6.981629392971246, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 8741 + }, + { + "epoch": 6.982428115015974, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 8742 + }, + { + "epoch": 6.983226837060703, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 8743 + }, + { + "epoch": 6.984025559105431, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 8744 + }, + { + "epoch": 6.98482428115016, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 8745 + }, + { + "epoch": 6.985623003194888, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 8746 + }, + { + "epoch": 6.986421725239617, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 8747 + }, + { + "epoch": 6.987220447284345, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 8748 + }, + { + "epoch": 6.988019169329074, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 8749 + }, + { + "epoch": 6.988817891373802, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 8750 + }, + { + "epoch": 6.98961661341853, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 8751 + }, + { + "epoch": 6.9904153354632586, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 8752 + }, + { + "epoch": 6.991214057507987, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 8753 + }, + { + "epoch": 6.992012779552716, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 8754 + }, + { + "epoch": 6.992811501597444, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 8755 + }, + { + "epoch": 6.993610223642173, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 8756 + }, + { + "epoch": 6.994408945686901, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 8757 + }, + { + "epoch": 6.99520766773163, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 8758 + }, + { + "epoch": 6.996006389776358, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 8759 + }, + { + "epoch": 6.996805111821086, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 8760 + }, + { + "epoch": 6.997603833865814, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 8761 + }, + { + "epoch": 6.998402555910543, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 8762 + }, + { + "epoch": 6.9992012779552715, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 8763 + }, + { + "epoch": 7.0, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 8764 + }, + { + "epoch": 7.0007987220447285, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 8765 + }, + { + "epoch": 7.001597444089457, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 8766 + }, + { + "epoch": 7.002396166134186, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1195, + "step": 8767 + }, + { + "epoch": 7.003194888178914, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 8768 + }, + { + "epoch": 7.003993610223642, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 8769 + }, + { + "epoch": 7.00479233226837, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 8770 + }, + { + "epoch": 7.005591054313099, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 8771 + }, + { + "epoch": 7.006389776357827, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 8772 + }, + { + "epoch": 7.007188498402556, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 8773 + }, + { + "epoch": 7.007987220447284, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 8774 + }, + { + "epoch": 7.008785942492013, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 8775 + }, + { + "epoch": 7.0095846645367414, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 8776 + }, + { + "epoch": 7.01038338658147, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 8777 + }, + { + "epoch": 7.0111821086261985, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1189, + "step": 8778 + }, + { + "epoch": 7.011980830670926, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 8779 + }, + { + "epoch": 7.012779552715655, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 8780 + }, + { + "epoch": 7.013578274760383, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 8781 + }, + { + "epoch": 7.014376996805112, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 8782 + }, + { + "epoch": 7.01517571884984, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 8783 + }, + { + "epoch": 7.015974440894569, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 8784 + }, + { + "epoch": 7.016773162939297, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 8785 + }, + { + "epoch": 7.017571884984026, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 8786 + }, + { + "epoch": 7.018370607028754, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 8787 + }, + { + "epoch": 7.019169329073482, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 8788 + }, + { + "epoch": 7.0199680511182105, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 8789 + }, + { + "epoch": 7.020766773162939, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 8790 + }, + { + "epoch": 7.021565495207668, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1157, + "step": 8791 + }, + { + "epoch": 7.022364217252396, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 8792 + }, + { + "epoch": 7.023162939297125, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 8793 + }, + { + "epoch": 7.023961661341853, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 8794 + }, + { + "epoch": 7.024760383386582, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 8795 + }, + { + "epoch": 7.02555910543131, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 8796 + }, + { + "epoch": 7.026357827476039, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 8797 + }, + { + "epoch": 7.027156549520766, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 8798 + }, + { + "epoch": 7.027955271565495, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 8799 + }, + { + "epoch": 7.0287539936102235, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 8800 + }, + { + "epoch": 7.029552715654952, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 8801 + }, + { + "epoch": 7.0303514376996805, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8802 + }, + { + "epoch": 7.031150159744409, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 8803 + }, + { + "epoch": 7.031948881789138, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 8804 + }, + { + "epoch": 7.032747603833866, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 8805 + }, + { + "epoch": 7.033546325878595, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 8806 + }, + { + "epoch": 7.034345047923322, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 8807 + }, + { + "epoch": 7.035143769968051, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 8808 + }, + { + "epoch": 7.035942492012779, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 8809 + }, + { + "epoch": 7.036741214057508, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8810 + }, + { + "epoch": 7.037539936102236, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 8811 + }, + { + "epoch": 7.038338658146965, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1187, + "step": 8812 + }, + { + "epoch": 7.039137380191693, + "grad_norm": 0.35546875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 8813 + }, + { + "epoch": 7.039936102236422, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 8814 + }, + { + "epoch": 7.0407348242811505, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 8815 + }, + { + "epoch": 7.041533546325879, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 8816 + }, + { + "epoch": 7.042332268370607, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 8817 + }, + { + "epoch": 7.043130990415335, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 8818 + }, + { + "epoch": 7.043929712460064, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 8819 + }, + { + "epoch": 7.044728434504792, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 8820 + }, + { + "epoch": 7.045527156549521, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 8821 + }, + { + "epoch": 7.046325878594249, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 8822 + }, + { + "epoch": 7.047124600638978, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 8823 + }, + { + "epoch": 7.047923322683706, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 8824 + }, + { + "epoch": 7.048722044728435, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 8825 + }, + { + "epoch": 7.0495207667731625, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 8826 + }, + { + "epoch": 7.050319488817891, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 8827 + }, + { + "epoch": 7.05111821086262, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 8828 + }, + { + "epoch": 7.051916932907348, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 8829 + }, + { + "epoch": 7.052715654952077, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 8830 + }, + { + "epoch": 7.053514376996805, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 8831 + }, + { + "epoch": 7.054313099041534, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 8832 + }, + { + "epoch": 7.055111821086262, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 8833 + }, + { + "epoch": 7.055910543130991, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 8834 + }, + { + "epoch": 7.056709265175719, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 8835 + }, + { + "epoch": 7.057507987220447, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 8836 + }, + { + "epoch": 7.0583067092651754, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 8837 + }, + { + "epoch": 7.059105431309904, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 8838 + }, + { + "epoch": 7.0599041533546325, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 8839 + }, + { + "epoch": 7.060702875399361, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 8840 + }, + { + "epoch": 7.06150159744409, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8841 + }, + { + "epoch": 7.062300319488818, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 8842 + }, + { + "epoch": 7.063099041533547, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 8843 + }, + { + "epoch": 7.063897763578275, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 8844 + }, + { + "epoch": 7.064696485623003, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 8845 + }, + { + "epoch": 7.065495207667731, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 8846 + }, + { + "epoch": 7.06629392971246, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 8847 + }, + { + "epoch": 7.067092651757188, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 8848 + }, + { + "epoch": 7.067891373801917, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 8849 + }, + { + "epoch": 7.068690095846645, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 8850 + }, + { + "epoch": 7.069488817891374, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 8851 + }, + { + "epoch": 7.0702875399361025, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 8852 + }, + { + "epoch": 7.071086261980831, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1219, + "step": 8853 + }, + { + "epoch": 7.0718849840255595, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 8854 + }, + { + "epoch": 7.072683706070287, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 8855 + }, + { + "epoch": 7.073482428115016, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 8856 + }, + { + "epoch": 7.074281150159744, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 8857 + }, + { + "epoch": 7.075079872204473, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 8858 + }, + { + "epoch": 7.075878594249201, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 8859 + }, + { + "epoch": 7.07667731629393, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 8860 + }, + { + "epoch": 7.077476038338658, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 8861 + }, + { + "epoch": 7.078274760383387, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 8862 + }, + { + "epoch": 7.079073482428115, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 8863 + }, + { + "epoch": 7.079872204472843, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 8864 + }, + { + "epoch": 7.080670926517572, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 8865 + }, + { + "epoch": 7.0814696485623, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 8866 + }, + { + "epoch": 7.082268370607029, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 8867 + }, + { + "epoch": 7.083067092651757, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 8868 + }, + { + "epoch": 7.083865814696486, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 8869 + }, + { + "epoch": 7.084664536741214, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 8870 + }, + { + "epoch": 7.085463258785943, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 8871 + }, + { + "epoch": 7.086261980830671, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 8872 + }, + { + "epoch": 7.0870607028754, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 8873 + }, + { + "epoch": 7.087859424920127, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 8874 + }, + { + "epoch": 7.088658146964856, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8875 + }, + { + "epoch": 7.0894568690095845, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 8876 + }, + { + "epoch": 7.090255591054313, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 8877 + }, + { + "epoch": 7.0910543130990416, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 8878 + }, + { + "epoch": 7.09185303514377, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 8879 + }, + { + "epoch": 7.092651757188499, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 8880 + }, + { + "epoch": 7.093450479233227, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 8881 + }, + { + "epoch": 7.094249201277956, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 8882 + }, + { + "epoch": 7.095047923322683, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 8883 + }, + { + "epoch": 7.095846645367412, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 8884 + }, + { + "epoch": 7.09664536741214, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 8885 + }, + { + "epoch": 7.097444089456869, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 8886 + }, + { + "epoch": 7.098242811501597, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 8887 + }, + { + "epoch": 7.099041533546326, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 8888 + }, + { + "epoch": 7.0998402555910545, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8889 + }, + { + "epoch": 7.100638977635783, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 8890 + }, + { + "epoch": 7.1014376996805115, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 8891 + }, + { + "epoch": 7.102236421725239, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 8892 + }, + { + "epoch": 7.103035143769968, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 8893 + }, + { + "epoch": 7.103833865814696, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 8894 + }, + { + "epoch": 7.104632587859425, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 8895 + }, + { + "epoch": 7.105431309904153, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 8896 + }, + { + "epoch": 7.106230031948882, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 8897 + }, + { + "epoch": 7.10702875399361, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 8898 + }, + { + "epoch": 7.107827476038339, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 8899 + }, + { + "epoch": 7.108626198083067, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 8900 + }, + { + "epoch": 7.109424920127796, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 8901 + }, + { + "epoch": 7.110223642172524, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 8902 + }, + { + "epoch": 7.111022364217252, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 8903 + }, + { + "epoch": 7.111821086261981, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 8904 + }, + { + "epoch": 7.112619808306709, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 8905 + }, + { + "epoch": 7.113418530351438, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 8906 + }, + { + "epoch": 7.114217252396166, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 8907 + }, + { + "epoch": 7.115015974440895, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 8908 + }, + { + "epoch": 7.115814696485623, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 8909 + }, + { + "epoch": 7.116613418530352, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 8910 + }, + { + "epoch": 7.11741214057508, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 8911 + }, + { + "epoch": 7.118210862619808, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 8912 + }, + { + "epoch": 7.1190095846645365, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 8913 + }, + { + "epoch": 7.119808306709265, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 8914 + }, + { + "epoch": 7.1206070287539935, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 8915 + }, + { + "epoch": 7.121405750798722, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 8916 + }, + { + "epoch": 7.122204472843451, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 8917 + }, + { + "epoch": 7.123003194888179, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 8918 + }, + { + "epoch": 7.123801916932908, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 8919 + }, + { + "epoch": 7.124600638977636, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 8920 + }, + { + "epoch": 7.125399361022364, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 8921 + }, + { + "epoch": 7.126198083067092, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 8922 + }, + { + "epoch": 7.126996805111821, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 8923 + }, + { + "epoch": 7.127795527156549, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 8924 + }, + { + "epoch": 7.128594249201278, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 8925 + }, + { + "epoch": 7.1293929712460065, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 8926 + }, + { + "epoch": 7.130191693290735, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 8927 + }, + { + "epoch": 7.1309904153354635, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 8928 + }, + { + "epoch": 7.131789137380192, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 8929 + }, + { + "epoch": 7.13258785942492, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 8930 + }, + { + "epoch": 7.133386581469648, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 8931 + }, + { + "epoch": 7.134185303514377, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 8932 + }, + { + "epoch": 7.134984025559105, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 8933 + }, + { + "epoch": 7.135782747603834, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 8934 + }, + { + "epoch": 7.136581469648562, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 8935 + }, + { + "epoch": 7.137380191693291, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 8936 + }, + { + "epoch": 7.138178913738019, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 8937 + }, + { + "epoch": 7.138977635782748, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 8938 + }, + { + "epoch": 7.139776357827476, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 8939 + }, + { + "epoch": 7.140575079872204, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 8940 + }, + { + "epoch": 7.141373801916933, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 8941 + }, + { + "epoch": 7.142172523961661, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 8942 + }, + { + "epoch": 7.14297124600639, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 8943 + }, + { + "epoch": 7.143769968051118, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 8944 + }, + { + "epoch": 7.144568690095847, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 8945 + }, + { + "epoch": 7.145367412140575, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 8946 + }, + { + "epoch": 7.146166134185304, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 8947 + }, + { + "epoch": 7.146964856230032, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 8948 + }, + { + "epoch": 7.147763578274761, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 8949 + }, + { + "epoch": 7.1485623003194885, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 8950 + }, + { + "epoch": 7.149361022364217, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 8951 + }, + { + "epoch": 7.1501597444089455, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 8952 + }, + { + "epoch": 7.150958466453674, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 8953 + }, + { + "epoch": 7.151757188498403, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 8954 + }, + { + "epoch": 7.152555910543131, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1167, + "step": 8955 + }, + { + "epoch": 7.15335463258786, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 8956 + }, + { + "epoch": 7.154153354632588, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 8957 + }, + { + "epoch": 7.154952076677317, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 8958 + }, + { + "epoch": 7.155750798722044, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 8959 + }, + { + "epoch": 7.156549520766773, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 8960 + }, + { + "epoch": 7.157348242811501, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 8961 + }, + { + "epoch": 7.15814696485623, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 8962 + }, + { + "epoch": 7.1589456869009584, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 8963 + }, + { + "epoch": 7.159744408945687, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 8964 + }, + { + "epoch": 7.1605431309904155, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 8965 + }, + { + "epoch": 7.161341853035144, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 8966 + }, + { + "epoch": 7.162140575079873, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 8967 + }, + { + "epoch": 7.1629392971246, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 8968 + }, + { + "epoch": 7.163738019169329, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 8969 + }, + { + "epoch": 7.164536741214057, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 8970 + }, + { + "epoch": 7.165335463258786, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 8971 + }, + { + "epoch": 7.166134185303514, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 8972 + }, + { + "epoch": 7.166932907348243, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 8973 + }, + { + "epoch": 7.167731629392971, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 8974 + }, + { + "epoch": 7.1685303514377, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 8975 + }, + { + "epoch": 7.169329073482428, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 8976 + }, + { + "epoch": 7.170127795527157, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 8977 + }, + { + "epoch": 7.170926517571885, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 8978 + }, + { + "epoch": 7.171725239616613, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 8979 + }, + { + "epoch": 7.172523961661342, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 8980 + }, + { + "epoch": 7.17332268370607, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 8981 + }, + { + "epoch": 7.174121405750799, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 8982 + }, + { + "epoch": 7.174920127795527, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 8983 + }, + { + "epoch": 7.175718849840256, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 8984 + }, + { + "epoch": 7.176517571884984, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 8985 + }, + { + "epoch": 7.177316293929713, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 8986 + }, + { + "epoch": 7.178115015974441, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 8987 + }, + { + "epoch": 7.178913738019169, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 8988 + }, + { + "epoch": 7.1797124600638975, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 8989 + }, + { + "epoch": 7.180511182108626, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 8990 + }, + { + "epoch": 7.181309904153355, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 8991 + }, + { + "epoch": 7.182108626198083, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 8992 + }, + { + "epoch": 7.182907348242812, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 8993 + }, + { + "epoch": 7.18370607028754, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 8994 + }, + { + "epoch": 7.184504792332269, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 8995 + }, + { + "epoch": 7.185303514376997, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 8996 + }, + { + "epoch": 7.186102236421725, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 8997 + }, + { + "epoch": 7.186900958466453, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 8998 + }, + { + "epoch": 7.187699680511182, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 8999 + }, + { + "epoch": 7.18849840255591, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 9000 + }, + { + "epoch": 7.189297124600639, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 9001 + }, + { + "epoch": 7.1900958466453675, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 9002 + }, + { + "epoch": 7.190894568690096, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 9003 + }, + { + "epoch": 7.1916932907348246, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 9004 + }, + { + "epoch": 7.192492012779553, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 9005 + }, + { + "epoch": 7.193290734824281, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 9006 + }, + { + "epoch": 7.194089456869009, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 9007 + }, + { + "epoch": 7.194888178913738, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 9008 + }, + { + "epoch": 7.195686900958466, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9009 + }, + { + "epoch": 7.196485623003195, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 9010 + }, + { + "epoch": 7.197284345047923, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 9011 + }, + { + "epoch": 7.198083067092652, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 9012 + }, + { + "epoch": 7.19888178913738, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 9013 + }, + { + "epoch": 7.199680511182109, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 9014 + }, + { + "epoch": 7.2004792332268375, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 9015 + }, + { + "epoch": 7.201277955271565, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 9016 + }, + { + "epoch": 7.202076677316294, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9017 + }, + { + "epoch": 7.202875399361022, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 9018 + }, + { + "epoch": 7.203674121405751, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 9019 + }, + { + "epoch": 7.204472843450479, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9020 + }, + { + "epoch": 7.205271565495208, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 9021 + }, + { + "epoch": 7.206070287539936, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 9022 + }, + { + "epoch": 7.206869009584665, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 9023 + }, + { + "epoch": 7.207667731629393, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 9024 + }, + { + "epoch": 7.208466453674121, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 9025 + }, + { + "epoch": 7.2092651757188495, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 9026 + }, + { + "epoch": 7.210063897763578, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 9027 + }, + { + "epoch": 7.210862619808307, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9028 + }, + { + "epoch": 7.211661341853035, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 9029 + }, + { + "epoch": 7.212460063897764, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9030 + }, + { + "epoch": 7.213258785942492, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 9031 + }, + { + "epoch": 7.214057507987221, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 9032 + }, + { + "epoch": 7.214856230031949, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 9033 + }, + { + "epoch": 7.215654952076678, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 9034 + }, + { + "epoch": 7.216453674121405, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 9035 + }, + { + "epoch": 7.217252396166134, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 9036 + }, + { + "epoch": 7.218051118210862, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 9037 + }, + { + "epoch": 7.218849840255591, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 9038 + }, + { + "epoch": 7.2196485623003195, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 9039 + }, + { + "epoch": 7.220447284345048, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 9040 + }, + { + "epoch": 7.2212460063897765, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9041 + }, + { + "epoch": 7.222044728434505, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 9042 + }, + { + "epoch": 7.222843450479234, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 9043 + }, + { + "epoch": 7.223642172523961, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9044 + }, + { + "epoch": 7.22444089456869, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9045 + }, + { + "epoch": 7.225239616613418, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9046 + }, + { + "epoch": 7.226038338658147, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 9047 + }, + { + "epoch": 7.226837060702875, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 9048 + }, + { + "epoch": 7.227635782747604, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9049 + }, + { + "epoch": 7.228434504792332, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 9050 + }, + { + "epoch": 7.229233226837061, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 9051 + }, + { + "epoch": 7.2300319488817895, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 9052 + }, + { + "epoch": 7.230830670926518, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 9053 + }, + { + "epoch": 7.231629392971246, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 9054 + }, + { + "epoch": 7.232428115015974, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 9055 + }, + { + "epoch": 7.233226837060703, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 9056 + }, + { + "epoch": 7.234025559105431, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 9057 + }, + { + "epoch": 7.23482428115016, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 9058 + }, + { + "epoch": 7.235623003194888, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 9059 + }, + { + "epoch": 7.236421725239617, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9060 + }, + { + "epoch": 7.237220447284345, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 9061 + }, + { + "epoch": 7.238019169329074, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 9062 + }, + { + "epoch": 7.2388178913738015, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 9063 + }, + { + "epoch": 7.23961661341853, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 9064 + }, + { + "epoch": 7.2404153354632586, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 9065 + }, + { + "epoch": 7.241214057507987, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9066 + }, + { + "epoch": 7.242012779552716, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 9067 + }, + { + "epoch": 7.242811501597444, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 9068 + }, + { + "epoch": 7.243610223642173, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 9069 + }, + { + "epoch": 7.244408945686901, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 9070 + }, + { + "epoch": 7.24520766773163, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 9071 + }, + { + "epoch": 7.246006389776358, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 9072 + }, + { + "epoch": 7.246805111821086, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 9073 + }, + { + "epoch": 7.247603833865814, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 9074 + }, + { + "epoch": 7.248402555910543, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 9075 + }, + { + "epoch": 7.2492012779552715, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 9076 + }, + { + "epoch": 7.25, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 9077 + }, + { + "epoch": 7.2507987220447285, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 9078 + }, + { + "epoch": 7.251597444089457, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 9079 + }, + { + "epoch": 7.252396166134186, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 9080 + }, + { + "epoch": 7.253194888178914, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 9081 + }, + { + "epoch": 7.253993610223642, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 9082 + }, + { + "epoch": 7.25479233226837, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 9083 + }, + { + "epoch": 7.255591054313099, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 9084 + }, + { + "epoch": 7.256389776357827, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 9085 + }, + { + "epoch": 7.257188498402556, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 9086 + }, + { + "epoch": 7.257987220447284, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 9087 + }, + { + "epoch": 7.258785942492013, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 9088 + }, + { + "epoch": 7.2595846645367414, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 9089 + }, + { + "epoch": 7.26038338658147, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 9090 + }, + { + "epoch": 7.261182108626198, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 9091 + }, + { + "epoch": 7.261980830670926, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 9092 + }, + { + "epoch": 7.262779552715655, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 9093 + }, + { + "epoch": 7.263578274760383, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 9094 + }, + { + "epoch": 7.264376996805112, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 9095 + }, + { + "epoch": 7.26517571884984, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9096 + }, + { + "epoch": 7.265974440894569, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 9097 + }, + { + "epoch": 7.266773162939297, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 9098 + }, + { + "epoch": 7.267571884984026, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 9099 + }, + { + "epoch": 7.268370607028754, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 9100 + }, + { + "epoch": 7.269169329073483, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9101 + }, + { + "epoch": 7.2699680511182105, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 9102 + }, + { + "epoch": 7.270766773162939, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 9103 + }, + { + "epoch": 7.271565495207668, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 9104 + }, + { + "epoch": 7.272364217252396, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 9105 + }, + { + "epoch": 7.273162939297125, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 9106 + }, + { + "epoch": 7.273961661341853, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 9107 + }, + { + "epoch": 7.274760383386582, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 9108 + }, + { + "epoch": 7.27555910543131, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 9109 + }, + { + "epoch": 7.276357827476039, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 9110 + }, + { + "epoch": 7.277156549520766, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 9111 + }, + { + "epoch": 7.277955271565495, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 9112 + }, + { + "epoch": 7.2787539936102235, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 9113 + }, + { + "epoch": 7.279552715654952, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 9114 + }, + { + "epoch": 7.2803514376996805, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1192, + "step": 9115 + }, + { + "epoch": 7.281150159744409, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 9116 + }, + { + "epoch": 7.281948881789138, + "grad_norm": 0.345703125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 9117 + }, + { + "epoch": 7.282747603833866, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 9118 + }, + { + "epoch": 7.283546325878595, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9119 + }, + { + "epoch": 7.284345047923322, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 9120 + }, + { + "epoch": 7.285143769968051, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 9121 + }, + { + "epoch": 7.285942492012779, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9122 + }, + { + "epoch": 7.286741214057508, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 9123 + }, + { + "epoch": 7.287539936102236, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 9124 + }, + { + "epoch": 7.288338658146965, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9125 + }, + { + "epoch": 7.289137380191693, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 9126 + }, + { + "epoch": 7.289936102236422, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 9127 + }, + { + "epoch": 7.2907348242811505, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 9128 + }, + { + "epoch": 7.291533546325878, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 9129 + }, + { + "epoch": 7.292332268370607, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 9130 + }, + { + "epoch": 7.293130990415335, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 9131 + }, + { + "epoch": 7.293929712460064, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 9132 + }, + { + "epoch": 7.294728434504792, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 9133 + }, + { + "epoch": 7.295527156549521, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 9134 + }, + { + "epoch": 7.296325878594249, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 9135 + }, + { + "epoch": 7.297124600638978, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 9136 + }, + { + "epoch": 7.297923322683706, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9137 + }, + { + "epoch": 7.298722044728435, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 9138 + }, + { + "epoch": 7.2995207667731625, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 9139 + }, + { + "epoch": 7.300319488817891, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9140 + }, + { + "epoch": 7.30111821086262, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 9141 + }, + { + "epoch": 7.301916932907348, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 9142 + }, + { + "epoch": 7.302715654952077, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 9143 + }, + { + "epoch": 7.303514376996805, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 9144 + }, + { + "epoch": 7.304313099041534, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 9145 + }, + { + "epoch": 7.305111821086262, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 9146 + }, + { + "epoch": 7.305910543130991, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 9147 + }, + { + "epoch": 7.306709265175719, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 9148 + }, + { + "epoch": 7.307507987220447, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 9149 + }, + { + "epoch": 7.3083067092651754, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 9150 + }, + { + "epoch": 7.309105431309904, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 9151 + }, + { + "epoch": 7.3099041533546325, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 9152 + }, + { + "epoch": 7.310702875399361, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 9153 + }, + { + "epoch": 7.31150159744409, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9154 + }, + { + "epoch": 7.312300319488818, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 9155 + }, + { + "epoch": 7.313099041533547, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 9156 + }, + { + "epoch": 7.313897763578275, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 9157 + }, + { + "epoch": 7.314696485623003, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 9158 + }, + { + "epoch": 7.315495207667731, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 9159 + }, + { + "epoch": 7.31629392971246, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 9160 + }, + { + "epoch": 7.317092651757188, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 9161 + }, + { + "epoch": 7.317891373801917, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 9162 + }, + { + "epoch": 7.318690095846645, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 9163 + }, + { + "epoch": 7.319488817891374, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 9164 + }, + { + "epoch": 7.3202875399361025, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 9165 + }, + { + "epoch": 7.321086261980831, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 9166 + }, + { + "epoch": 7.321884984025559, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9167 + }, + { + "epoch": 7.322683706070287, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 9168 + }, + { + "epoch": 7.323482428115016, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 9169 + }, + { + "epoch": 7.324281150159744, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 9170 + }, + { + "epoch": 7.325079872204473, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 9171 + }, + { + "epoch": 7.325878594249201, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 9172 + }, + { + "epoch": 7.32667731629393, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9173 + }, + { + "epoch": 7.327476038338658, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 9174 + }, + { + "epoch": 7.328274760383387, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 9175 + }, + { + "epoch": 7.329073482428115, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 9176 + }, + { + "epoch": 7.329872204472843, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 9177 + }, + { + "epoch": 7.330670926517572, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 9178 + }, + { + "epoch": 7.3314696485623, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 9179 + }, + { + "epoch": 7.332268370607029, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 9180 + }, + { + "epoch": 7.333067092651757, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 9181 + }, + { + "epoch": 7.333865814696486, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 9182 + }, + { + "epoch": 7.334664536741214, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9183 + }, + { + "epoch": 7.335463258785943, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 9184 + }, + { + "epoch": 7.336261980830671, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 9185 + }, + { + "epoch": 7.3370607028754, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 9186 + }, + { + "epoch": 7.337859424920127, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9187 + }, + { + "epoch": 7.338658146964856, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9188 + }, + { + "epoch": 7.3394568690095845, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 9189 + }, + { + "epoch": 7.340255591054313, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 9190 + }, + { + "epoch": 7.3410543130990416, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 9191 + }, + { + "epoch": 7.34185303514377, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 9192 + }, + { + "epoch": 7.342651757188499, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 9193 + }, + { + "epoch": 7.343450479233227, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 9194 + }, + { + "epoch": 7.344249201277956, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 9195 + }, + { + "epoch": 7.345047923322683, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9196 + }, + { + "epoch": 7.345846645367412, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 9197 + }, + { + "epoch": 7.34664536741214, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 9198 + }, + { + "epoch": 7.347444089456869, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 9199 + }, + { + "epoch": 7.348242811501597, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 9200 + }, + { + "epoch": 7.349041533546326, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 9201 + }, + { + "epoch": 7.3498402555910545, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 9202 + }, + { + "epoch": 7.350638977635783, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9203 + }, + { + "epoch": 7.3514376996805115, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 9204 + }, + { + "epoch": 7.352236421725239, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 9205 + }, + { + "epoch": 7.353035143769968, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9206 + }, + { + "epoch": 7.353833865814696, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 9207 + }, + { + "epoch": 7.354632587859425, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 9208 + }, + { + "epoch": 7.355431309904153, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 9209 + }, + { + "epoch": 7.356230031948882, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 9210 + }, + { + "epoch": 7.35702875399361, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 9211 + }, + { + "epoch": 7.357827476038339, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 9212 + }, + { + "epoch": 7.358626198083067, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1181, + "step": 9213 + }, + { + "epoch": 7.359424920127796, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9214 + }, + { + "epoch": 7.360223642172524, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 9215 + }, + { + "epoch": 7.361022364217252, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 9216 + }, + { + "epoch": 7.361821086261981, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 9217 + }, + { + "epoch": 7.362619808306709, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 9218 + }, + { + "epoch": 7.363418530351438, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 9219 + }, + { + "epoch": 7.364217252396166, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 9220 + }, + { + "epoch": 7.365015974440895, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9221 + }, + { + "epoch": 7.365814696485623, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 9222 + }, + { + "epoch": 7.366613418530352, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 9223 + }, + { + "epoch": 7.36741214057508, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 9224 + }, + { + "epoch": 7.368210862619808, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 9225 + }, + { + "epoch": 7.3690095846645365, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 9226 + }, + { + "epoch": 7.369808306709265, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 9227 + }, + { + "epoch": 7.3706070287539935, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 9228 + }, + { + "epoch": 7.371405750798722, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 9229 + }, + { + "epoch": 7.372204472843451, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 9230 + }, + { + "epoch": 7.373003194888179, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 9231 + }, + { + "epoch": 7.373801916932908, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 9232 + }, + { + "epoch": 7.374600638977636, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 9233 + }, + { + "epoch": 7.375399361022364, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 9234 + }, + { + "epoch": 7.376198083067092, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 9235 + }, + { + "epoch": 7.376996805111821, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9236 + }, + { + "epoch": 7.377795527156549, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 9237 + }, + { + "epoch": 7.378594249201278, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 9238 + }, + { + "epoch": 7.3793929712460065, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9239 + }, + { + "epoch": 7.380191693290735, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 9240 + }, + { + "epoch": 7.3809904153354635, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 9241 + }, + { + "epoch": 7.381789137380192, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 9242 + }, + { + "epoch": 7.38258785942492, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 9243 + }, + { + "epoch": 7.383386581469648, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 9244 + }, + { + "epoch": 7.384185303514377, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9245 + }, + { + "epoch": 7.384984025559105, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 9246 + }, + { + "epoch": 7.385782747603834, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 9247 + }, + { + "epoch": 7.386581469648562, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 9248 + }, + { + "epoch": 7.387380191693291, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 9249 + }, + { + "epoch": 7.388178913738019, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 9250 + }, + { + "epoch": 7.388977635782748, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 9251 + }, + { + "epoch": 7.389776357827476, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 9252 + }, + { + "epoch": 7.390575079872204, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 9253 + }, + { + "epoch": 7.391373801916933, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 9254 + }, + { + "epoch": 7.392172523961661, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 9255 + }, + { + "epoch": 7.39297124600639, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 9256 + }, + { + "epoch": 7.393769968051118, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 9257 + }, + { + "epoch": 7.394568690095847, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 9258 + }, + { + "epoch": 7.395367412140575, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 9259 + }, + { + "epoch": 7.396166134185304, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 9260 + }, + { + "epoch": 7.396964856230032, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 9261 + }, + { + "epoch": 7.397763578274761, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 9262 + }, + { + "epoch": 7.3985623003194885, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 9263 + }, + { + "epoch": 7.399361022364217, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 9264 + }, + { + "epoch": 7.4001597444089455, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 9265 + }, + { + "epoch": 7.400958466453674, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 9266 + }, + { + "epoch": 7.401757188498403, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 9267 + }, + { + "epoch": 7.402555910543131, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 9268 + }, + { + "epoch": 7.40335463258786, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 9269 + }, + { + "epoch": 7.404153354632588, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 9270 + }, + { + "epoch": 7.404952076677317, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 9271 + }, + { + "epoch": 7.405750798722044, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 9272 + }, + { + "epoch": 7.406549520766773, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9273 + }, + { + "epoch": 7.407348242811501, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 9274 + }, + { + "epoch": 7.40814696485623, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 9275 + }, + { + "epoch": 7.4089456869009584, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 9276 + }, + { + "epoch": 7.409744408945687, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 9277 + }, + { + "epoch": 7.4105431309904155, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 9278 + }, + { + "epoch": 7.411341853035144, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 9279 + }, + { + "epoch": 7.412140575079873, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9280 + }, + { + "epoch": 7.4129392971246, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 9281 + }, + { + "epoch": 7.413738019169329, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9282 + }, + { + "epoch": 7.414536741214057, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 9283 + }, + { + "epoch": 7.415335463258786, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9284 + }, + { + "epoch": 7.416134185303514, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9285 + }, + { + "epoch": 7.416932907348243, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 9286 + }, + { + "epoch": 7.417731629392971, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 9287 + }, + { + "epoch": 7.4185303514377, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 9288 + }, + { + "epoch": 7.419329073482428, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 9289 + }, + { + "epoch": 7.420127795527157, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 9290 + }, + { + "epoch": 7.420926517571885, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 9291 + }, + { + "epoch": 7.421725239616613, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 9292 + }, + { + "epoch": 7.422523961661342, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 9293 + }, + { + "epoch": 7.42332268370607, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 9294 + }, + { + "epoch": 7.424121405750799, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 9295 + }, + { + "epoch": 7.424920127795527, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 9296 + }, + { + "epoch": 7.425718849840256, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 9297 + }, + { + "epoch": 7.426517571884984, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 9298 + }, + { + "epoch": 7.427316293929713, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 9299 + }, + { + "epoch": 7.428115015974441, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 9300 + }, + { + "epoch": 7.428913738019169, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 9301 + }, + { + "epoch": 7.4297124600638975, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 9302 + }, + { + "epoch": 7.430511182108626, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 9303 + }, + { + "epoch": 7.431309904153355, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 9304 + }, + { + "epoch": 7.432108626198083, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9305 + }, + { + "epoch": 7.432907348242812, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 9306 + }, + { + "epoch": 7.43370607028754, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9307 + }, + { + "epoch": 7.434504792332269, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9308 + }, + { + "epoch": 7.435303514376997, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9309 + }, + { + "epoch": 7.436102236421725, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 9310 + }, + { + "epoch": 7.436900958466453, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 9311 + }, + { + "epoch": 7.437699680511182, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 9312 + }, + { + "epoch": 7.43849840255591, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 9313 + }, + { + "epoch": 7.439297124600639, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 9314 + }, + { + "epoch": 7.4400958466453675, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 9315 + }, + { + "epoch": 7.440894568690096, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 9316 + }, + { + "epoch": 7.4416932907348246, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 9317 + }, + { + "epoch": 7.442492012779553, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 9318 + }, + { + "epoch": 7.443290734824281, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 9319 + }, + { + "epoch": 7.444089456869009, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 9320 + }, + { + "epoch": 7.444888178913738, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9321 + }, + { + "epoch": 7.445686900958466, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 9322 + }, + { + "epoch": 7.446485623003195, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 9323 + }, + { + "epoch": 7.447284345047923, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 9324 + }, + { + "epoch": 7.448083067092652, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 9325 + }, + { + "epoch": 7.44888178913738, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 9326 + }, + { + "epoch": 7.449680511182109, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 9327 + }, + { + "epoch": 7.4504792332268375, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9328 + }, + { + "epoch": 7.451277955271565, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 9329 + }, + { + "epoch": 7.452076677316294, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 9330 + }, + { + "epoch": 7.452875399361022, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 9331 + }, + { + "epoch": 7.453674121405751, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 9332 + }, + { + "epoch": 7.454472843450479, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 9333 + }, + { + "epoch": 7.455271565495208, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 9334 + }, + { + "epoch": 7.456070287539936, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 9335 + }, + { + "epoch": 7.456869009584665, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 9336 + }, + { + "epoch": 7.457667731629393, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 9337 + }, + { + "epoch": 7.458466453674122, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 9338 + }, + { + "epoch": 7.4592651757188495, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 9339 + }, + { + "epoch": 7.460063897763578, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9340 + }, + { + "epoch": 7.460862619808307, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 9341 + }, + { + "epoch": 7.461661341853035, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 9342 + }, + { + "epoch": 7.462460063897764, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 9343 + }, + { + "epoch": 7.463258785942492, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 9344 + }, + { + "epoch": 7.464057507987221, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 9345 + }, + { + "epoch": 7.464856230031949, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9346 + }, + { + "epoch": 7.465654952076678, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 9347 + }, + { + "epoch": 7.466453674121405, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 9348 + }, + { + "epoch": 7.467252396166134, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 9349 + }, + { + "epoch": 7.468051118210862, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 9350 + }, + { + "epoch": 7.468849840255591, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9351 + }, + { + "epoch": 7.4696485623003195, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 9352 + }, + { + "epoch": 7.470447284345048, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 9353 + }, + { + "epoch": 7.4712460063897765, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 9354 + }, + { + "epoch": 7.472044728434505, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 9355 + }, + { + "epoch": 7.472843450479234, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 9356 + }, + { + "epoch": 7.473642172523961, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9357 + }, + { + "epoch": 7.47444089456869, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 9358 + }, + { + "epoch": 7.475239616613418, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 9359 + }, + { + "epoch": 7.476038338658147, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 9360 + }, + { + "epoch": 7.476837060702875, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 9361 + }, + { + "epoch": 7.477635782747604, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 9362 + }, + { + "epoch": 7.478434504792332, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 9363 + }, + { + "epoch": 7.479233226837061, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 9364 + }, + { + "epoch": 7.4800319488817895, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9365 + }, + { + "epoch": 7.480830670926517, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 9366 + }, + { + "epoch": 7.481629392971246, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 9367 + }, + { + "epoch": 7.482428115015974, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 9368 + }, + { + "epoch": 7.483226837060703, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 9369 + }, + { + "epoch": 7.484025559105431, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 9370 + }, + { + "epoch": 7.48482428115016, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 9371 + }, + { + "epoch": 7.485623003194888, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 9372 + }, + { + "epoch": 7.486421725239617, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 9373 + }, + { + "epoch": 7.487220447284345, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 9374 + }, + { + "epoch": 7.488019169329074, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 9375 + }, + { + "epoch": 7.488817891373802, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 9376 + }, + { + "epoch": 7.48961661341853, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 9377 + }, + { + "epoch": 7.4904153354632586, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 9378 + }, + { + "epoch": 7.491214057507987, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 9379 + }, + { + "epoch": 7.492012779552716, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 9380 + }, + { + "epoch": 7.492811501597444, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 9381 + }, + { + "epoch": 7.493610223642173, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 9382 + }, + { + "epoch": 7.494408945686901, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9383 + }, + { + "epoch": 7.49520766773163, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 9384 + }, + { + "epoch": 7.496006389776358, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 9385 + }, + { + "epoch": 7.496805111821086, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9386 + }, + { + "epoch": 7.497603833865814, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 9387 + }, + { + "epoch": 7.498402555910543, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 9388 + }, + { + "epoch": 7.4992012779552715, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 9389 + }, + { + "epoch": 7.5, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 9390 + }, + { + "epoch": 7.5007987220447285, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 9391 + }, + { + "epoch": 7.501597444089457, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 9392 + }, + { + "epoch": 7.502396166134186, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 9393 + }, + { + "epoch": 7.503194888178914, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 9394 + }, + { + "epoch": 7.503993610223642, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 9395 + }, + { + "epoch": 7.50479233226837, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 9396 + }, + { + "epoch": 7.505591054313099, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 9397 + }, + { + "epoch": 7.506389776357827, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 9398 + }, + { + "epoch": 7.507188498402556, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 9399 + }, + { + "epoch": 7.507987220447284, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 9400 + }, + { + "epoch": 7.508785942492013, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9401 + }, + { + "epoch": 7.5095846645367414, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 9402 + }, + { + "epoch": 7.51038338658147, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 9403 + }, + { + "epoch": 7.511182108626198, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 9404 + }, + { + "epoch": 7.511980830670926, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 9405 + }, + { + "epoch": 7.512779552715655, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9406 + }, + { + "epoch": 7.513578274760383, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 9407 + }, + { + "epoch": 7.514376996805112, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 9408 + }, + { + "epoch": 7.51517571884984, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 9409 + }, + { + "epoch": 7.515974440894569, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 9410 + }, + { + "epoch": 7.516773162939297, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9411 + }, + { + "epoch": 7.517571884984026, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 9412 + }, + { + "epoch": 7.518370607028754, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9413 + }, + { + "epoch": 7.519169329073483, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 9414 + }, + { + "epoch": 7.5199680511182105, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 9415 + }, + { + "epoch": 7.520766773162939, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9416 + }, + { + "epoch": 7.521565495207668, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 9417 + }, + { + "epoch": 7.522364217252396, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 9418 + }, + { + "epoch": 7.523162939297125, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 9419 + }, + { + "epoch": 7.523961661341853, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 9420 + }, + { + "epoch": 7.524760383386582, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 9421 + }, + { + "epoch": 7.52555910543131, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 9422 + }, + { + "epoch": 7.526357827476039, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 9423 + }, + { + "epoch": 7.527156549520766, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 9424 + }, + { + "epoch": 7.527955271565495, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 9425 + }, + { + "epoch": 7.5287539936102235, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 9426 + }, + { + "epoch": 7.529552715654952, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1188, + "step": 9427 + }, + { + "epoch": 7.5303514376996805, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 9428 + }, + { + "epoch": 7.531150159744409, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 9429 + }, + { + "epoch": 7.531948881789138, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9430 + }, + { + "epoch": 7.532747603833866, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9431 + }, + { + "epoch": 7.533546325878595, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 9432 + }, + { + "epoch": 7.534345047923322, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 9433 + }, + { + "epoch": 7.535143769968051, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 9434 + }, + { + "epoch": 7.535942492012779, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 9435 + }, + { + "epoch": 7.536741214057508, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 9436 + }, + { + "epoch": 7.537539936102236, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 9437 + }, + { + "epoch": 7.538338658146965, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9438 + }, + { + "epoch": 7.539137380191693, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9439 + }, + { + "epoch": 7.539936102236422, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 9440 + }, + { + "epoch": 7.5407348242811505, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 9441 + }, + { + "epoch": 7.541533546325878, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 9442 + }, + { + "epoch": 7.542332268370607, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 9443 + }, + { + "epoch": 7.543130990415335, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 9444 + }, + { + "epoch": 7.543929712460064, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 9445 + }, + { + "epoch": 7.544728434504792, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 9446 + }, + { + "epoch": 7.545527156549521, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 9447 + }, + { + "epoch": 7.546325878594249, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 9448 + }, + { + "epoch": 7.547124600638978, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9449 + }, + { + "epoch": 7.547923322683706, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 9450 + }, + { + "epoch": 7.548722044728435, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 9451 + }, + { + "epoch": 7.549520766773163, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 9452 + }, + { + "epoch": 7.550319488817891, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 9453 + }, + { + "epoch": 7.55111821086262, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 9454 + }, + { + "epoch": 7.551916932907348, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 9455 + }, + { + "epoch": 7.552715654952077, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 9456 + }, + { + "epoch": 7.553514376996805, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 9457 + }, + { + "epoch": 7.554313099041534, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 9458 + }, + { + "epoch": 7.555111821086262, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 9459 + }, + { + "epoch": 7.555910543130991, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 9460 + }, + { + "epoch": 7.556709265175719, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 9461 + }, + { + "epoch": 7.557507987220447, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9462 + }, + { + "epoch": 7.5583067092651754, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9463 + }, + { + "epoch": 7.559105431309904, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 9464 + }, + { + "epoch": 7.5599041533546325, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 9465 + }, + { + "epoch": 7.560702875399361, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 9466 + }, + { + "epoch": 7.56150159744409, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 9467 + }, + { + "epoch": 7.562300319488818, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 9468 + }, + { + "epoch": 7.563099041533547, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 9469 + }, + { + "epoch": 7.563897763578275, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 9470 + }, + { + "epoch": 7.564696485623003, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 9471 + }, + { + "epoch": 7.565495207667731, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 9472 + }, + { + "epoch": 7.56629392971246, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 9473 + }, + { + "epoch": 7.567092651757188, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1164, + "step": 9474 + }, + { + "epoch": 7.567891373801917, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 9475 + }, + { + "epoch": 7.568690095846645, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 9476 + }, + { + "epoch": 7.569488817891374, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 9477 + }, + { + "epoch": 7.5702875399361025, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 9478 + }, + { + "epoch": 7.571086261980831, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9479 + }, + { + "epoch": 7.571884984025559, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9480 + }, + { + "epoch": 7.572683706070287, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 9481 + }, + { + "epoch": 7.573482428115016, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 9482 + }, + { + "epoch": 7.574281150159744, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 9483 + }, + { + "epoch": 7.575079872204473, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 9484 + }, + { + "epoch": 7.575878594249201, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 9485 + }, + { + "epoch": 7.57667731629393, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 9486 + }, + { + "epoch": 7.577476038338658, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 9487 + }, + { + "epoch": 7.578274760383387, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 9488 + }, + { + "epoch": 7.5790734824281145, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 9489 + }, + { + "epoch": 7.579872204472844, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1207, + "step": 9490 + }, + { + "epoch": 7.580670926517572, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 9491 + }, + { + "epoch": 7.5814696485623, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 9492 + }, + { + "epoch": 7.582268370607029, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9493 + }, + { + "epoch": 7.583067092651757, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 9494 + }, + { + "epoch": 7.583865814696486, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9495 + }, + { + "epoch": 7.584664536741214, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9496 + }, + { + "epoch": 7.585463258785943, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9497 + }, + { + "epoch": 7.586261980830671, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 9498 + }, + { + "epoch": 7.5870607028754, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 9499 + }, + { + "epoch": 7.587859424920127, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 9500 + }, + { + "epoch": 7.588658146964856, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 9501 + }, + { + "epoch": 7.5894568690095845, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 9502 + }, + { + "epoch": 7.590255591054313, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 9503 + }, + { + "epoch": 7.5910543130990416, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 9504 + }, + { + "epoch": 7.59185303514377, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 9505 + }, + { + "epoch": 7.592651757188499, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 9506 + }, + { + "epoch": 7.593450479233227, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9507 + }, + { + "epoch": 7.594249201277956, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 9508 + }, + { + "epoch": 7.595047923322683, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 9509 + }, + { + "epoch": 7.595846645367412, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 9510 + }, + { + "epoch": 7.59664536741214, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 9511 + }, + { + "epoch": 7.597444089456869, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 9512 + }, + { + "epoch": 7.598242811501597, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 9513 + }, + { + "epoch": 7.599041533546326, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9514 + }, + { + "epoch": 7.5998402555910545, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 9515 + }, + { + "epoch": 7.600638977635783, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 9516 + }, + { + "epoch": 7.6014376996805115, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 9517 + }, + { + "epoch": 7.602236421725239, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 9518 + }, + { + "epoch": 7.603035143769968, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 9519 + }, + { + "epoch": 7.603833865814696, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 9520 + }, + { + "epoch": 7.604632587859425, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 9521 + }, + { + "epoch": 7.605431309904153, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 9522 + }, + { + "epoch": 7.606230031948882, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 9523 + }, + { + "epoch": 7.60702875399361, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 9524 + }, + { + "epoch": 7.607827476038339, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 9525 + }, + { + "epoch": 7.608626198083067, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9526 + }, + { + "epoch": 7.609424920127795, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 9527 + }, + { + "epoch": 7.6102236421725244, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 9528 + }, + { + "epoch": 7.611022364217252, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 9529 + }, + { + "epoch": 7.611821086261981, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 9530 + }, + { + "epoch": 7.612619808306709, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 9531 + }, + { + "epoch": 7.613418530351438, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 9532 + }, + { + "epoch": 7.614217252396166, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 9533 + }, + { + "epoch": 7.615015974440895, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 9534 + }, + { + "epoch": 7.615814696485623, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 9535 + }, + { + "epoch": 7.616613418530352, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 9536 + }, + { + "epoch": 7.61741214057508, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 9537 + }, + { + "epoch": 7.618210862619808, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 9538 + }, + { + "epoch": 7.6190095846645365, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9539 + }, + { + "epoch": 7.619808306709265, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 9540 + }, + { + "epoch": 7.6206070287539935, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 9541 + }, + { + "epoch": 7.621405750798722, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 9542 + }, + { + "epoch": 7.622204472843451, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 9543 + }, + { + "epoch": 7.623003194888179, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 9544 + }, + { + "epoch": 7.623801916932908, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 9545 + }, + { + "epoch": 7.624600638977636, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 9546 + }, + { + "epoch": 7.625399361022364, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 9547 + }, + { + "epoch": 7.626198083067092, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 9548 + }, + { + "epoch": 7.626996805111821, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 9549 + }, + { + "epoch": 7.627795527156549, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 9550 + }, + { + "epoch": 7.628594249201278, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9551 + }, + { + "epoch": 7.6293929712460065, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 9552 + }, + { + "epoch": 7.630191693290735, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 9553 + }, + { + "epoch": 7.6309904153354635, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9554 + }, + { + "epoch": 7.631789137380192, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 9555 + }, + { + "epoch": 7.63258785942492, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 9556 + }, + { + "epoch": 7.633386581469648, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 9557 + }, + { + "epoch": 7.634185303514377, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9558 + }, + { + "epoch": 7.634984025559105, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 9559 + }, + { + "epoch": 7.635782747603834, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 9560 + }, + { + "epoch": 7.636581469648562, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 9561 + }, + { + "epoch": 7.637380191693291, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 9562 + }, + { + "epoch": 7.638178913738019, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 9563 + }, + { + "epoch": 7.638977635782748, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 9564 + }, + { + "epoch": 7.6397763578274756, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 9565 + }, + { + "epoch": 7.640575079872205, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 9566 + }, + { + "epoch": 7.641373801916933, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9567 + }, + { + "epoch": 7.642172523961661, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 9568 + }, + { + "epoch": 7.64297124600639, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 9569 + }, + { + "epoch": 7.643769968051118, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 9570 + }, + { + "epoch": 7.644568690095847, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 9571 + }, + { + "epoch": 7.645367412140575, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 9572 + }, + { + "epoch": 7.646166134185304, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 9573 + }, + { + "epoch": 7.646964856230032, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 9574 + }, + { + "epoch": 7.647763578274761, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 9575 + }, + { + "epoch": 7.6485623003194885, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 9576 + }, + { + "epoch": 7.649361022364217, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9577 + }, + { + "epoch": 7.6501597444089455, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 9578 + }, + { + "epoch": 7.650958466453674, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 9579 + }, + { + "epoch": 7.651757188498403, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 9580 + }, + { + "epoch": 7.652555910543131, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 9581 + }, + { + "epoch": 7.65335463258786, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 9582 + }, + { + "epoch": 7.654153354632588, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 9583 + }, + { + "epoch": 7.654952076677317, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 9584 + }, + { + "epoch": 7.655750798722044, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 9585 + }, + { + "epoch": 7.656549520766773, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 9586 + }, + { + "epoch": 7.657348242811501, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 9587 + }, + { + "epoch": 7.65814696485623, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9588 + }, + { + "epoch": 7.6589456869009584, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 9589 + }, + { + "epoch": 7.659744408945687, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 9590 + }, + { + "epoch": 7.6605431309904155, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 9591 + }, + { + "epoch": 7.661341853035144, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 9592 + }, + { + "epoch": 7.662140575079873, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 9593 + }, + { + "epoch": 7.6629392971246, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 9594 + }, + { + "epoch": 7.663738019169329, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9595 + }, + { + "epoch": 7.664536741214057, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 9596 + }, + { + "epoch": 7.665335463258786, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 9597 + }, + { + "epoch": 7.666134185303514, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 9598 + }, + { + "epoch": 7.666932907348243, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 9599 + }, + { + "epoch": 7.667731629392971, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 9600 + }, + { + "epoch": 7.6685303514377, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 9601 + }, + { + "epoch": 7.669329073482428, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 9602 + }, + { + "epoch": 7.670127795527156, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 9603 + }, + { + "epoch": 7.6709265175718855, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 9604 + }, + { + "epoch": 7.671725239616613, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 9605 + }, + { + "epoch": 7.672523961661342, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 9606 + }, + { + "epoch": 7.67332268370607, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 9607 + }, + { + "epoch": 7.674121405750799, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 9608 + }, + { + "epoch": 7.674920127795527, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9609 + }, + { + "epoch": 7.675718849840256, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 9610 + }, + { + "epoch": 7.676517571884984, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 9611 + }, + { + "epoch": 7.677316293929713, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 9612 + }, + { + "epoch": 7.678115015974441, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 9613 + }, + { + "epoch": 7.678913738019169, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 9614 + }, + { + "epoch": 7.6797124600638975, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9615 + }, + { + "epoch": 7.680511182108626, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 9616 + }, + { + "epoch": 7.681309904153355, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 9617 + }, + { + "epoch": 7.682108626198083, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 9618 + }, + { + "epoch": 7.682907348242812, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 9619 + }, + { + "epoch": 7.68370607028754, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 9620 + }, + { + "epoch": 7.684504792332269, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 9621 + }, + { + "epoch": 7.685303514376997, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 9622 + }, + { + "epoch": 7.686102236421725, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 9623 + }, + { + "epoch": 7.686900958466453, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 9624 + }, + { + "epoch": 7.687699680511182, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 9625 + }, + { + "epoch": 7.68849840255591, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 9626 + }, + { + "epoch": 7.689297124600639, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 9627 + }, + { + "epoch": 7.6900958466453675, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 9628 + }, + { + "epoch": 7.690894568690096, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9629 + }, + { + "epoch": 7.6916932907348246, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 9630 + }, + { + "epoch": 7.692492012779553, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 9631 + }, + { + "epoch": 7.693290734824281, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9632 + }, + { + "epoch": 7.694089456869009, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 9633 + }, + { + "epoch": 7.694888178913738, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9634 + }, + { + "epoch": 7.695686900958466, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9635 + }, + { + "epoch": 7.696485623003195, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 9636 + }, + { + "epoch": 7.697284345047923, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 9637 + }, + { + "epoch": 7.698083067092652, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 9638 + }, + { + "epoch": 7.69888178913738, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 9639 + }, + { + "epoch": 7.699680511182109, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 9640 + }, + { + "epoch": 7.700479233226837, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 9641 + }, + { + "epoch": 7.701277955271565, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 9642 + }, + { + "epoch": 7.702076677316294, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9643 + }, + { + "epoch": 7.702875399361022, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 9644 + }, + { + "epoch": 7.703674121405751, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 9645 + }, + { + "epoch": 7.704472843450479, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 9646 + }, + { + "epoch": 7.705271565495208, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 9647 + }, + { + "epoch": 7.706070287539936, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 9648 + }, + { + "epoch": 7.706869009584665, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 9649 + }, + { + "epoch": 7.707667731629393, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 9650 + }, + { + "epoch": 7.708466453674122, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 9651 + }, + { + "epoch": 7.7092651757188495, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 9652 + }, + { + "epoch": 7.710063897763578, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 9653 + }, + { + "epoch": 7.710862619808307, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 9654 + }, + { + "epoch": 7.711661341853035, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 9655 + }, + { + "epoch": 7.712460063897764, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1148, + "step": 9656 + }, + { + "epoch": 7.713258785942492, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9657 + }, + { + "epoch": 7.714057507987221, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 9658 + }, + { + "epoch": 7.714856230031949, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 9659 + }, + { + "epoch": 7.715654952076678, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 9660 + }, + { + "epoch": 7.716453674121405, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 9661 + }, + { + "epoch": 7.717252396166134, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 9662 + }, + { + "epoch": 7.718051118210862, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 9663 + }, + { + "epoch": 7.718849840255591, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 9664 + }, + { + "epoch": 7.7196485623003195, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 9665 + }, + { + "epoch": 7.720447284345048, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 9666 + }, + { + "epoch": 7.7212460063897765, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 9667 + }, + { + "epoch": 7.722044728434505, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 9668 + }, + { + "epoch": 7.722843450479234, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 9669 + }, + { + "epoch": 7.723642172523961, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 9670 + }, + { + "epoch": 7.72444089456869, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 9671 + }, + { + "epoch": 7.725239616613418, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 9672 + }, + { + "epoch": 7.726038338658147, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 9673 + }, + { + "epoch": 7.726837060702875, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 9674 + }, + { + "epoch": 7.727635782747604, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 9675 + }, + { + "epoch": 7.728434504792332, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 9676 + }, + { + "epoch": 7.729233226837061, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9677 + }, + { + "epoch": 7.7300319488817895, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 9678 + }, + { + "epoch": 7.730830670926517, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 9679 + }, + { + "epoch": 7.731629392971246, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 9680 + }, + { + "epoch": 7.732428115015974, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 9681 + }, + { + "epoch": 7.733226837060703, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 9682 + }, + { + "epoch": 7.734025559105431, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 9683 + }, + { + "epoch": 7.73482428115016, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 9684 + }, + { + "epoch": 7.735623003194888, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 9685 + }, + { + "epoch": 7.736421725239617, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9686 + }, + { + "epoch": 7.737220447284345, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 9687 + }, + { + "epoch": 7.738019169329074, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9688 + }, + { + "epoch": 7.738817891373802, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 9689 + }, + { + "epoch": 7.73961661341853, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 9690 + }, + { + "epoch": 7.7404153354632586, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 9691 + }, + { + "epoch": 7.741214057507987, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 9692 + }, + { + "epoch": 7.742012779552716, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 9693 + }, + { + "epoch": 7.742811501597444, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 9694 + }, + { + "epoch": 7.743610223642173, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 9695 + }, + { + "epoch": 7.744408945686901, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 9696 + }, + { + "epoch": 7.74520766773163, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 9697 + }, + { + "epoch": 7.746006389776358, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 9698 + }, + { + "epoch": 7.746805111821086, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 9699 + }, + { + "epoch": 7.747603833865814, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 9700 + }, + { + "epoch": 7.748402555910543, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 9701 + }, + { + "epoch": 7.7492012779552715, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 9702 + }, + { + "epoch": 7.75, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 9703 + }, + { + "epoch": 7.7507987220447285, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1194, + "step": 9704 + }, + { + "epoch": 7.751597444089457, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 9705 + }, + { + "epoch": 7.752396166134186, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 9706 + }, + { + "epoch": 7.753194888178914, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 9707 + }, + { + "epoch": 7.753993610223642, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 9708 + }, + { + "epoch": 7.75479233226837, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 9709 + }, + { + "epoch": 7.755591054313099, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 9710 + }, + { + "epoch": 7.756389776357827, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 9711 + }, + { + "epoch": 7.757188498402556, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 9712 + }, + { + "epoch": 7.757987220447284, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 9713 + }, + { + "epoch": 7.758785942492013, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1206, + "step": 9714 + }, + { + "epoch": 7.7595846645367414, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 9715 + }, + { + "epoch": 7.76038338658147, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 9716 + }, + { + "epoch": 7.761182108626198, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 9717 + }, + { + "epoch": 7.761980830670926, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 9718 + }, + { + "epoch": 7.762779552715655, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 9719 + }, + { + "epoch": 7.763578274760383, + "grad_norm": 0.404296875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 9720 + }, + { + "epoch": 7.764376996805112, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 9721 + }, + { + "epoch": 7.76517571884984, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 9722 + }, + { + "epoch": 7.765974440894569, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9723 + }, + { + "epoch": 7.766773162939297, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 9724 + }, + { + "epoch": 7.767571884984026, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 9725 + }, + { + "epoch": 7.768370607028754, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 9726 + }, + { + "epoch": 7.769169329073483, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 9727 + }, + { + "epoch": 7.7699680511182105, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 9728 + }, + { + "epoch": 7.770766773162939, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 9729 + }, + { + "epoch": 7.771565495207668, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 9730 + }, + { + "epoch": 7.772364217252396, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 9731 + }, + { + "epoch": 7.773162939297125, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 9732 + }, + { + "epoch": 7.773961661341853, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 9733 + }, + { + "epoch": 7.774760383386582, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 9734 + }, + { + "epoch": 7.77555910543131, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 9735 + }, + { + "epoch": 7.776357827476039, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 9736 + }, + { + "epoch": 7.777156549520766, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 9737 + }, + { + "epoch": 7.777955271565495, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 9738 + }, + { + "epoch": 7.7787539936102235, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 9739 + }, + { + "epoch": 7.779552715654952, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 9740 + }, + { + "epoch": 7.7803514376996805, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9741 + }, + { + "epoch": 7.781150159744409, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9742 + }, + { + "epoch": 7.781948881789138, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 9743 + }, + { + "epoch": 7.782747603833866, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 9744 + }, + { + "epoch": 7.783546325878595, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 9745 + }, + { + "epoch": 7.784345047923322, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 9746 + }, + { + "epoch": 7.785143769968051, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 9747 + }, + { + "epoch": 7.785942492012779, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 9748 + }, + { + "epoch": 7.786741214057508, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 9749 + }, + { + "epoch": 7.787539936102236, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 9750 + }, + { + "epoch": 7.788338658146965, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 9751 + }, + { + "epoch": 7.789137380191693, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 9752 + }, + { + "epoch": 7.789936102236422, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 9753 + }, + { + "epoch": 7.7907348242811505, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 9754 + }, + { + "epoch": 7.791533546325878, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 9755 + }, + { + "epoch": 7.792332268370607, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 9756 + }, + { + "epoch": 7.793130990415335, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9757 + }, + { + "epoch": 7.793929712460064, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 9758 + }, + { + "epoch": 7.794728434504792, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 9759 + }, + { + "epoch": 7.795527156549521, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 9760 + }, + { + "epoch": 7.796325878594249, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9761 + }, + { + "epoch": 7.797124600638978, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 9762 + }, + { + "epoch": 7.797923322683706, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 9763 + }, + { + "epoch": 7.798722044728435, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 9764 + }, + { + "epoch": 7.799520766773163, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 9765 + }, + { + "epoch": 7.800319488817891, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 9766 + }, + { + "epoch": 7.80111821086262, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 9767 + }, + { + "epoch": 7.801916932907348, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 9768 + }, + { + "epoch": 7.802715654952077, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 9769 + }, + { + "epoch": 7.803514376996805, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 9770 + }, + { + "epoch": 7.804313099041534, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9771 + }, + { + "epoch": 7.805111821086262, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 9772 + }, + { + "epoch": 7.805910543130991, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 9773 + }, + { + "epoch": 7.806709265175719, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 9774 + }, + { + "epoch": 7.807507987220447, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 9775 + }, + { + "epoch": 7.8083067092651754, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 9776 + }, + { + "epoch": 7.809105431309904, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 9777 + }, + { + "epoch": 7.8099041533546325, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 9778 + }, + { + "epoch": 7.810702875399361, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 9779 + }, + { + "epoch": 7.81150159744409, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 9780 + }, + { + "epoch": 7.812300319488818, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 9781 + }, + { + "epoch": 7.813099041533547, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 9782 + }, + { + "epoch": 7.813897763578275, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9783 + }, + { + "epoch": 7.814696485623003, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9784 + }, + { + "epoch": 7.815495207667731, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 9785 + }, + { + "epoch": 7.81629392971246, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9786 + }, + { + "epoch": 7.817092651757188, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9787 + }, + { + "epoch": 7.817891373801917, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 9788 + }, + { + "epoch": 7.818690095846645, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 9789 + }, + { + "epoch": 7.819488817891374, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 9790 + }, + { + "epoch": 7.8202875399361025, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 9791 + }, + { + "epoch": 7.821086261980831, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 9792 + }, + { + "epoch": 7.821884984025559, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 9793 + }, + { + "epoch": 7.822683706070287, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 9794 + }, + { + "epoch": 7.823482428115016, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 9795 + }, + { + "epoch": 7.824281150159744, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9796 + }, + { + "epoch": 7.825079872204473, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 9797 + }, + { + "epoch": 7.825878594249201, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 9798 + }, + { + "epoch": 7.82667731629393, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 9799 + }, + { + "epoch": 7.827476038338658, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 9800 + }, + { + "epoch": 7.828274760383387, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 9801 + }, + { + "epoch": 7.8290734824281145, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 9802 + }, + { + "epoch": 7.829872204472844, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 9803 + }, + { + "epoch": 7.830670926517572, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 9804 + }, + { + "epoch": 7.8314696485623, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 9805 + }, + { + "epoch": 7.832268370607029, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 9806 + }, + { + "epoch": 7.833067092651757, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 9807 + }, + { + "epoch": 7.833865814696486, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 9808 + }, + { + "epoch": 7.834664536741214, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 9809 + }, + { + "epoch": 7.835463258785943, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 9810 + }, + { + "epoch": 7.836261980830671, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9811 + }, + { + "epoch": 7.8370607028754, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 9812 + }, + { + "epoch": 7.837859424920127, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 9813 + }, + { + "epoch": 7.838658146964856, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 9814 + }, + { + "epoch": 7.8394568690095845, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 9815 + }, + { + "epoch": 7.840255591054313, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 9816 + }, + { + "epoch": 7.8410543130990416, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9817 + }, + { + "epoch": 7.84185303514377, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 9818 + }, + { + "epoch": 7.842651757188499, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 9819 + }, + { + "epoch": 7.843450479233227, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 9820 + }, + { + "epoch": 7.844249201277956, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 9821 + }, + { + "epoch": 7.845047923322683, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 9822 + }, + { + "epoch": 7.845846645367412, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 9823 + }, + { + "epoch": 7.84664536741214, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 9824 + }, + { + "epoch": 7.847444089456869, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 9825 + }, + { + "epoch": 7.848242811501597, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 9826 + }, + { + "epoch": 7.849041533546326, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 9827 + }, + { + "epoch": 7.8498402555910545, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9828 + }, + { + "epoch": 7.850638977635783, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 9829 + }, + { + "epoch": 7.8514376996805115, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 9830 + }, + { + "epoch": 7.852236421725239, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 9831 + }, + { + "epoch": 7.853035143769968, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 9832 + }, + { + "epoch": 7.853833865814696, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 9833 + }, + { + "epoch": 7.854632587859425, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 9834 + }, + { + "epoch": 7.855431309904153, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 9835 + }, + { + "epoch": 7.856230031948882, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 9836 + }, + { + "epoch": 7.85702875399361, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 9837 + }, + { + "epoch": 7.857827476038339, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 9838 + }, + { + "epoch": 7.858626198083067, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 9839 + }, + { + "epoch": 7.859424920127795, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 9840 + }, + { + "epoch": 7.8602236421725244, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 9841 + }, + { + "epoch": 7.861022364217252, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 9842 + }, + { + "epoch": 7.861821086261981, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 9843 + }, + { + "epoch": 7.862619808306709, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9844 + }, + { + "epoch": 7.863418530351438, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 9845 + }, + { + "epoch": 7.864217252396166, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 9846 + }, + { + "epoch": 7.865015974440895, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 9847 + }, + { + "epoch": 7.865814696485623, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 9848 + }, + { + "epoch": 7.866613418530352, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 9849 + }, + { + "epoch": 7.86741214057508, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 9850 + }, + { + "epoch": 7.868210862619808, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9851 + }, + { + "epoch": 7.8690095846645365, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 9852 + }, + { + "epoch": 7.869808306709265, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 9853 + }, + { + "epoch": 7.8706070287539935, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 9854 + }, + { + "epoch": 7.871405750798722, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1215, + "step": 9855 + }, + { + "epoch": 7.872204472843451, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 9856 + }, + { + "epoch": 7.873003194888179, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 9857 + }, + { + "epoch": 7.873801916932908, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 9858 + }, + { + "epoch": 7.874600638977636, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 9859 + }, + { + "epoch": 7.875399361022364, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 9860 + }, + { + "epoch": 7.876198083067092, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 9861 + }, + { + "epoch": 7.876996805111821, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 9862 + }, + { + "epoch": 7.877795527156549, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 9863 + }, + { + "epoch": 7.878594249201278, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 9864 + }, + { + "epoch": 7.8793929712460065, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 9865 + }, + { + "epoch": 7.880191693290735, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9866 + }, + { + "epoch": 7.8809904153354635, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 9867 + }, + { + "epoch": 7.881789137380192, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 9868 + }, + { + "epoch": 7.88258785942492, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 9869 + }, + { + "epoch": 7.883386581469648, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 9870 + }, + { + "epoch": 7.884185303514377, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 9871 + }, + { + "epoch": 7.884984025559105, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 9872 + }, + { + "epoch": 7.885782747603834, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 9873 + }, + { + "epoch": 7.886581469648562, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 9874 + }, + { + "epoch": 7.887380191693291, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 9875 + }, + { + "epoch": 7.888178913738019, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 9876 + }, + { + "epoch": 7.888977635782748, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 9877 + }, + { + "epoch": 7.8897763578274756, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 9878 + }, + { + "epoch": 7.890575079872205, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 9879 + }, + { + "epoch": 7.891373801916933, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 9880 + }, + { + "epoch": 7.892172523961661, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 9881 + }, + { + "epoch": 7.89297124600639, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 9882 + }, + { + "epoch": 7.893769968051118, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 9883 + }, + { + "epoch": 7.894568690095847, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 9884 + }, + { + "epoch": 7.895367412140575, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1201, + "step": 9885 + }, + { + "epoch": 7.896166134185304, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 9886 + }, + { + "epoch": 7.896964856230032, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 9887 + }, + { + "epoch": 7.897763578274761, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 9888 + }, + { + "epoch": 7.8985623003194885, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 9889 + }, + { + "epoch": 7.899361022364217, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 9890 + }, + { + "epoch": 7.9001597444089455, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 9891 + }, + { + "epoch": 7.900958466453674, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 9892 + }, + { + "epoch": 7.901757188498403, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 9893 + }, + { + "epoch": 7.902555910543131, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 9894 + }, + { + "epoch": 7.90335463258786, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 9895 + }, + { + "epoch": 7.904153354632588, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9896 + }, + { + "epoch": 7.904952076677317, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 9897 + }, + { + "epoch": 7.905750798722044, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 9898 + }, + { + "epoch": 7.906549520766773, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 9899 + }, + { + "epoch": 7.907348242811501, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9900 + }, + { + "epoch": 7.90814696485623, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 9901 + }, + { + "epoch": 7.9089456869009584, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 9902 + }, + { + "epoch": 7.909744408945687, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 9903 + }, + { + "epoch": 7.9105431309904155, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 9904 + }, + { + "epoch": 7.911341853035144, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 9905 + }, + { + "epoch": 7.912140575079873, + "grad_norm": 0.365234375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 9906 + }, + { + "epoch": 7.9129392971246, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 9907 + }, + { + "epoch": 7.913738019169329, + "grad_norm": 0.380859375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 9908 + }, + { + "epoch": 7.914536741214057, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 9909 + }, + { + "epoch": 7.915335463258786, + "grad_norm": 0.345703125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 9910 + }, + { + "epoch": 7.916134185303514, + "grad_norm": 0.49609375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9911 + }, + { + "epoch": 7.916932907348243, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 9912 + }, + { + "epoch": 7.917731629392971, + "grad_norm": 0.453125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 9913 + }, + { + "epoch": 7.9185303514377, + "grad_norm": 0.322265625, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 9914 + }, + { + "epoch": 7.919329073482428, + "grad_norm": 0.470703125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 9915 + }, + { + "epoch": 7.920127795527156, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 9916 + }, + { + "epoch": 7.9209265175718855, + "grad_norm": 0.6015625, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 9917 + }, + { + "epoch": 7.921725239616613, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 9918 + }, + { + "epoch": 7.922523961661342, + "grad_norm": 0.48046875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 9919 + }, + { + "epoch": 7.92332268370607, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 9920 + }, + { + "epoch": 7.924121405750799, + "grad_norm": 0.55078125, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 9921 + }, + { + "epoch": 7.924920127795527, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 9922 + }, + { + "epoch": 7.925718849840256, + "grad_norm": 0.447265625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 9923 + }, + { + "epoch": 7.926517571884984, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 9924 + }, + { + "epoch": 7.927316293929713, + "grad_norm": 0.5078125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 9925 + }, + { + "epoch": 7.928115015974441, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 9926 + }, + { + "epoch": 7.928913738019169, + "grad_norm": 0.625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 9927 + }, + { + "epoch": 7.9297124600638975, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 9928 + }, + { + "epoch": 7.930511182108626, + "grad_norm": 0.498046875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 9929 + }, + { + "epoch": 7.931309904153355, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 9930 + }, + { + "epoch": 7.932108626198083, + "grad_norm": 0.43359375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 9931 + }, + { + "epoch": 7.932907348242812, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 9932 + }, + { + "epoch": 7.93370607028754, + "grad_norm": 0.515625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 9933 + }, + { + "epoch": 7.934504792332269, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 9934 + }, + { + "epoch": 7.935303514376997, + "grad_norm": 0.5234375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 9935 + }, + { + "epoch": 7.936102236421725, + "grad_norm": 0.388671875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 9936 + }, + { + "epoch": 7.936900958466453, + "grad_norm": 1.0234375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 9937 + }, + { + "epoch": 7.937699680511182, + "grad_norm": 0.49609375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 9938 + }, + { + "epoch": 7.93849840255591, + "grad_norm": 0.484375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 9939 + }, + { + "epoch": 7.939297124600639, + "grad_norm": 0.7734375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 9940 + }, + { + "epoch": 7.9400958466453675, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 9941 + }, + { + "epoch": 7.940894568690096, + "grad_norm": 0.53515625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 9942 + }, + { + "epoch": 7.9416932907348246, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 9943 + }, + { + "epoch": 7.942492012779553, + "grad_norm": 0.640625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 9944 + }, + { + "epoch": 7.943290734824281, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 9945 + }, + { + "epoch": 7.944089456869009, + "grad_norm": 0.79296875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 9946 + }, + { + "epoch": 7.944888178913738, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9947 + }, + { + "epoch": 7.945686900958466, + "grad_norm": 0.8046875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 9948 + }, + { + "epoch": 7.946485623003195, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 9949 + }, + { + "epoch": 7.947284345047923, + "grad_norm": 1.09375, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 9950 + }, + { + "epoch": 7.948083067092652, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9951 + }, + { + "epoch": 7.94888178913738, + "grad_norm": 0.80859375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 9952 + }, + { + "epoch": 7.949680511182109, + "grad_norm": 0.404296875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 9953 + }, + { + "epoch": 7.950479233226837, + "grad_norm": 0.546875, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 9954 + }, + { + "epoch": 7.951277955271565, + "grad_norm": 0.44921875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 9955 + }, + { + "epoch": 7.952076677316294, + "grad_norm": 0.439453125, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 9956 + }, + { + "epoch": 7.952875399361022, + "grad_norm": 0.5859375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 9957 + }, + { + "epoch": 7.953674121405751, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 9958 + }, + { + "epoch": 7.954472843450479, + "grad_norm": 0.58984375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 9959 + }, + { + "epoch": 7.955271565495208, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 9960 + }, + { + "epoch": 7.956070287539936, + "grad_norm": 0.62109375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 9961 + }, + { + "epoch": 7.956869009584665, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 9962 + }, + { + "epoch": 7.957667731629393, + "grad_norm": 0.79296875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 9963 + }, + { + "epoch": 7.958466453674122, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 9964 + }, + { + "epoch": 7.9592651757188495, + "grad_norm": 0.9296875, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 9965 + }, + { + "epoch": 7.960063897763578, + "grad_norm": 0.5, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9966 + }, + { + "epoch": 7.960862619808307, + "grad_norm": 0.58203125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 9967 + }, + { + "epoch": 7.961661341853035, + "grad_norm": 0.640625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 9968 + }, + { + "epoch": 7.962460063897764, + "grad_norm": 0.470703125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 9969 + }, + { + "epoch": 7.963258785942492, + "grad_norm": 1.03125, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 9970 + }, + { + "epoch": 7.964057507987221, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 9971 + }, + { + "epoch": 7.964856230031949, + "grad_norm": 0.58984375, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 9972 + }, + { + "epoch": 7.965654952076678, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 9973 + }, + { + "epoch": 7.966453674121405, + "grad_norm": 0.75, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 9974 + }, + { + "epoch": 7.967252396166134, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 9975 + }, + { + "epoch": 7.968051118210862, + "grad_norm": 0.58984375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 9976 + }, + { + "epoch": 7.968849840255591, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 9977 + }, + { + "epoch": 7.9696485623003195, + "grad_norm": 0.76171875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 9978 + }, + { + "epoch": 7.970447284345048, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 9979 + }, + { + "epoch": 7.9712460063897765, + "grad_norm": 0.50390625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 9980 + }, + { + "epoch": 7.972044728434505, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 9981 + }, + { + "epoch": 7.972843450479234, + "grad_norm": 0.416015625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 9982 + }, + { + "epoch": 7.973642172523961, + "grad_norm": 0.388671875, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 9983 + }, + { + "epoch": 7.97444089456869, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 9984 + }, + { + "epoch": 7.975239616613418, + "grad_norm": 0.416015625, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 9985 + }, + { + "epoch": 7.976038338658147, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 9986 + }, + { + "epoch": 7.976837060702875, + "grad_norm": 0.3828125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 9987 + }, + { + "epoch": 7.977635782747604, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 9988 + }, + { + "epoch": 7.978434504792332, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 9989 + }, + { + "epoch": 7.979233226837061, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 9990 + }, + { + "epoch": 7.9800319488817895, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 9991 + }, + { + "epoch": 7.980830670926517, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 9992 + }, + { + "epoch": 7.981629392971246, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 9993 + }, + { + "epoch": 7.982428115015974, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 9994 + }, + { + "epoch": 7.983226837060703, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 9995 + }, + { + "epoch": 7.984025559105431, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 9996 + }, + { + "epoch": 7.98482428115016, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 9997 + }, + { + "epoch": 7.985623003194888, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 9998 + }, + { + "epoch": 7.986421725239617, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 9999 + }, + { + "epoch": 7.987220447284345, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 10000 + }, + { + "epoch": 7.988019169329074, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 10001 + }, + { + "epoch": 7.988817891373802, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 10002 + }, + { + "epoch": 7.98961661341853, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10003 + }, + { + "epoch": 7.9904153354632586, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 10004 + }, + { + "epoch": 7.991214057507987, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 10005 + }, + { + "epoch": 7.992012779552716, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 10006 + }, + { + "epoch": 7.992811501597444, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 10007 + }, + { + "epoch": 7.993610223642173, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 10008 + }, + { + "epoch": 7.994408945686901, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 10009 + }, + { + "epoch": 7.99520766773163, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 10010 + }, + { + "epoch": 7.996006389776358, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 10011 + }, + { + "epoch": 7.996805111821086, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 10012 + }, + { + "epoch": 7.997603833865814, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 10013 + }, + { + "epoch": 7.998402555910543, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 10014 + }, + { + "epoch": 7.9992012779552715, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 10015 + }, + { + "epoch": 8.0, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 10016 + }, + { + "epoch": 8.000798722044728, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 10017 + }, + { + "epoch": 8.001597444089457, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 10018 + }, + { + "epoch": 8.002396166134185, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 10019 + }, + { + "epoch": 8.003194888178914, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 10020 + }, + { + "epoch": 8.003993610223642, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 10021 + }, + { + "epoch": 8.004792332268371, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 10022 + }, + { + "epoch": 8.005591054313099, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 10023 + }, + { + "epoch": 8.006389776357828, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 10024 + }, + { + "epoch": 8.007188498402556, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 10025 + }, + { + "epoch": 8.007987220447284, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10026 + }, + { + "epoch": 8.008785942492013, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 10027 + }, + { + "epoch": 8.00958466453674, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 10028 + }, + { + "epoch": 8.01038338658147, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 10029 + }, + { + "epoch": 8.011182108626198, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 10030 + }, + { + "epoch": 8.011980830670927, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1169, + "step": 10031 + }, + { + "epoch": 8.012779552715655, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 10032 + }, + { + "epoch": 8.013578274760384, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 10033 + }, + { + "epoch": 8.014376996805112, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 10034 + }, + { + "epoch": 8.01517571884984, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 10035 + }, + { + "epoch": 8.015974440894569, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 10036 + }, + { + "epoch": 8.016773162939296, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 10037 + }, + { + "epoch": 8.017571884984026, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 10038 + }, + { + "epoch": 8.018370607028753, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10039 + }, + { + "epoch": 8.019169329073483, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 10040 + }, + { + "epoch": 8.01996805111821, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 10041 + }, + { + "epoch": 8.02076677316294, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 10042 + }, + { + "epoch": 8.021565495207668, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 10043 + }, + { + "epoch": 8.022364217252397, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 10044 + }, + { + "epoch": 8.023162939297125, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 10045 + }, + { + "epoch": 8.023961661341852, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 10046 + }, + { + "epoch": 8.024760383386582, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 10047 + }, + { + "epoch": 8.02555910543131, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 10048 + }, + { + "epoch": 8.026357827476039, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 10049 + }, + { + "epoch": 8.027156549520766, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 10050 + }, + { + "epoch": 8.027955271565496, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 10051 + }, + { + "epoch": 8.028753993610223, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 10052 + }, + { + "epoch": 8.029552715654953, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 10053 + }, + { + "epoch": 8.03035143769968, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 10054 + }, + { + "epoch": 8.031150159744408, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 10055 + }, + { + "epoch": 8.031948881789138, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 10056 + }, + { + "epoch": 8.032747603833865, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 10057 + }, + { + "epoch": 8.033546325878595, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 10058 + }, + { + "epoch": 8.034345047923322, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 10059 + }, + { + "epoch": 8.035143769968052, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10060 + }, + { + "epoch": 8.03594249201278, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 10061 + }, + { + "epoch": 8.036741214057509, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 10062 + }, + { + "epoch": 8.037539936102236, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 10063 + }, + { + "epoch": 8.038338658146964, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 10064 + }, + { + "epoch": 8.039137380191693, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 10065 + }, + { + "epoch": 8.039936102236421, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 10066 + }, + { + "epoch": 8.04073482428115, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 10067 + }, + { + "epoch": 8.041533546325878, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 10068 + }, + { + "epoch": 8.042332268370608, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 10069 + }, + { + "epoch": 8.043130990415335, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 10070 + }, + { + "epoch": 8.043929712460065, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 10071 + }, + { + "epoch": 8.044728434504792, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 10072 + }, + { + "epoch": 8.04552715654952, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 10073 + }, + { + "epoch": 8.04632587859425, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 10074 + }, + { + "epoch": 8.047124600638977, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 10075 + }, + { + "epoch": 8.047923322683706, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 10076 + }, + { + "epoch": 8.048722044728434, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 10077 + }, + { + "epoch": 8.049520766773163, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 10078 + }, + { + "epoch": 8.050319488817891, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 10079 + }, + { + "epoch": 8.05111821086262, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 10080 + }, + { + "epoch": 8.051916932907348, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 10081 + }, + { + "epoch": 8.052715654952078, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 10082 + }, + { + "epoch": 8.053514376996805, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 10083 + }, + { + "epoch": 8.054313099041533, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 10084 + }, + { + "epoch": 8.055111821086262, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 10085 + }, + { + "epoch": 8.05591054313099, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 10086 + }, + { + "epoch": 8.05670926517572, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 10087 + }, + { + "epoch": 8.057507987220447, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 10088 + }, + { + "epoch": 8.058306709265176, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 10089 + }, + { + "epoch": 8.059105431309904, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 10090 + }, + { + "epoch": 8.059904153354633, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10091 + }, + { + "epoch": 8.060702875399361, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 10092 + }, + { + "epoch": 8.061501597444089, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 10093 + }, + { + "epoch": 8.062300319488818, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 10094 + }, + { + "epoch": 8.063099041533546, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 10095 + }, + { + "epoch": 8.063897763578275, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 10096 + }, + { + "epoch": 8.064696485623003, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 10097 + }, + { + "epoch": 8.065495207667732, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10098 + }, + { + "epoch": 8.06629392971246, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 10099 + }, + { + "epoch": 8.06709265175719, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 10100 + }, + { + "epoch": 8.067891373801917, + "grad_norm": 0.09423828125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 10101 + }, + { + "epoch": 8.068690095846645, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 10102 + }, + { + "epoch": 8.069488817891374, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10103 + }, + { + "epoch": 8.070287539936102, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 10104 + }, + { + "epoch": 8.071086261980831, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 10105 + }, + { + "epoch": 8.071884984025559, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 10106 + }, + { + "epoch": 8.072683706070288, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 10107 + }, + { + "epoch": 8.073482428115016, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 10108 + }, + { + "epoch": 8.074281150159745, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 10109 + }, + { + "epoch": 8.075079872204473, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 10110 + }, + { + "epoch": 8.0758785942492, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 10111 + }, + { + "epoch": 8.07667731629393, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10112 + }, + { + "epoch": 8.077476038338657, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 10113 + }, + { + "epoch": 8.078274760383387, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 10114 + }, + { + "epoch": 8.079073482428115, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 10115 + }, + { + "epoch": 8.079872204472844, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 10116 + }, + { + "epoch": 8.080670926517572, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 10117 + }, + { + "epoch": 8.081469648562301, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 10118 + }, + { + "epoch": 8.082268370607029, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 10119 + }, + { + "epoch": 8.083067092651758, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 10120 + }, + { + "epoch": 8.083865814696486, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 10121 + }, + { + "epoch": 8.084664536741213, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 10122 + }, + { + "epoch": 8.085463258785943, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 10123 + }, + { + "epoch": 8.08626198083067, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 10124 + }, + { + "epoch": 8.0870607028754, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 10125 + }, + { + "epoch": 8.087859424920127, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 10126 + }, + { + "epoch": 8.088658146964857, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 10127 + }, + { + "epoch": 8.089456869009584, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 10128 + }, + { + "epoch": 8.090255591054314, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 10129 + }, + { + "epoch": 8.091054313099042, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 10130 + }, + { + "epoch": 8.09185303514377, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 10131 + }, + { + "epoch": 8.092651757188499, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 10132 + }, + { + "epoch": 8.093450479233226, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 10133 + }, + { + "epoch": 8.094249201277956, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 10134 + }, + { + "epoch": 8.095047923322683, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 10135 + }, + { + "epoch": 8.095846645367413, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 10136 + }, + { + "epoch": 8.09664536741214, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 10137 + }, + { + "epoch": 8.09744408945687, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 10138 + }, + { + "epoch": 8.098242811501597, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 10139 + }, + { + "epoch": 8.099041533546325, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 10140 + }, + { + "epoch": 8.099840255591054, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 10141 + }, + { + "epoch": 8.100638977635782, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 10142 + }, + { + "epoch": 8.101437699680512, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 10143 + }, + { + "epoch": 8.10223642172524, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 10144 + }, + { + "epoch": 8.103035143769969, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 10145 + }, + { + "epoch": 8.103833865814696, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 10146 + }, + { + "epoch": 8.104632587859426, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 10147 + }, + { + "epoch": 8.105431309904153, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 10148 + }, + { + "epoch": 8.106230031948881, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 10149 + }, + { + "epoch": 8.10702875399361, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 10150 + }, + { + "epoch": 8.107827476038338, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 10151 + }, + { + "epoch": 8.108626198083067, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 10152 + }, + { + "epoch": 8.109424920127795, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 10153 + }, + { + "epoch": 8.110223642172524, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 10154 + }, + { + "epoch": 8.111022364217252, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 10155 + }, + { + "epoch": 8.111821086261982, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 10156 + }, + { + "epoch": 8.11261980830671, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 10157 + }, + { + "epoch": 8.113418530351439, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 10158 + }, + { + "epoch": 8.114217252396166, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 10159 + }, + { + "epoch": 8.115015974440894, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 10160 + }, + { + "epoch": 8.115814696485623, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 10161 + }, + { + "epoch": 8.116613418530351, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 10162 + }, + { + "epoch": 8.11741214057508, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1237, + "step": 10163 + }, + { + "epoch": 8.118210862619808, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 10164 + }, + { + "epoch": 8.119009584664537, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10165 + }, + { + "epoch": 8.119808306709265, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 10166 + }, + { + "epoch": 8.120607028753994, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 10167 + }, + { + "epoch": 8.121405750798722, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 10168 + }, + { + "epoch": 8.12220447284345, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 10169 + }, + { + "epoch": 8.12300319488818, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 10170 + }, + { + "epoch": 8.123801916932907, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 10171 + }, + { + "epoch": 8.124600638977636, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 10172 + }, + { + "epoch": 8.125399361022364, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 10173 + }, + { + "epoch": 8.126198083067093, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 10174 + }, + { + "epoch": 8.12699680511182, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 10175 + }, + { + "epoch": 8.12779552715655, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 10176 + }, + { + "epoch": 8.128594249201278, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 10177 + }, + { + "epoch": 8.129392971246006, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 10178 + }, + { + "epoch": 8.130191693290735, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 10179 + }, + { + "epoch": 8.130990415335463, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 10180 + }, + { + "epoch": 8.131789137380192, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 10181 + }, + { + "epoch": 8.13258785942492, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 10182 + }, + { + "epoch": 8.133386581469649, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 10183 + }, + { + "epoch": 8.134185303514377, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 10184 + }, + { + "epoch": 8.134984025559106, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 10185 + }, + { + "epoch": 8.135782747603834, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 10186 + }, + { + "epoch": 8.136581469648561, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 10187 + }, + { + "epoch": 8.13738019169329, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 10188 + }, + { + "epoch": 8.138178913738018, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 10189 + }, + { + "epoch": 8.138977635782748, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 10190 + }, + { + "epoch": 8.139776357827476, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 10191 + }, + { + "epoch": 8.140575079872205, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 10192 + }, + { + "epoch": 8.141373801916933, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 10193 + }, + { + "epoch": 8.142172523961662, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10194 + }, + { + "epoch": 8.14297124600639, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 10195 + }, + { + "epoch": 8.143769968051119, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10196 + }, + { + "epoch": 8.144568690095847, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 10197 + }, + { + "epoch": 8.145367412140574, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10198 + }, + { + "epoch": 8.146166134185304, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 10199 + }, + { + "epoch": 8.146964856230031, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 10200 + }, + { + "epoch": 8.14776357827476, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 10201 + }, + { + "epoch": 8.148562300319488, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 10202 + }, + { + "epoch": 8.149361022364218, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 10203 + }, + { + "epoch": 8.150159744408946, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 10204 + }, + { + "epoch": 8.150958466453675, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 10205 + }, + { + "epoch": 8.151757188498403, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 10206 + }, + { + "epoch": 8.15255591054313, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 10207 + }, + { + "epoch": 8.15335463258786, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 10208 + }, + { + "epoch": 8.154153354632587, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 10209 + }, + { + "epoch": 8.154952076677317, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 10210 + }, + { + "epoch": 8.155750798722044, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 10211 + }, + { + "epoch": 8.156549520766774, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 10212 + }, + { + "epoch": 8.157348242811501, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 10213 + }, + { + "epoch": 8.15814696485623, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10214 + }, + { + "epoch": 8.158945686900958, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 10215 + }, + { + "epoch": 8.159744408945686, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10216 + }, + { + "epoch": 8.160543130990416, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 10217 + }, + { + "epoch": 8.161341853035143, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 10218 + }, + { + "epoch": 8.162140575079873, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 10219 + }, + { + "epoch": 8.1629392971246, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 10220 + }, + { + "epoch": 8.16373801916933, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 10221 + }, + { + "epoch": 8.164536741214057, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 10222 + }, + { + "epoch": 8.165335463258787, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 10223 + }, + { + "epoch": 8.166134185303514, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 10224 + }, + { + "epoch": 8.166932907348242, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 10225 + }, + { + "epoch": 8.167731629392971, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 10226 + }, + { + "epoch": 8.168530351437699, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10227 + }, + { + "epoch": 8.169329073482428, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 10228 + }, + { + "epoch": 8.170127795527156, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 10229 + }, + { + "epoch": 8.170926517571885, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 10230 + }, + { + "epoch": 8.171725239616613, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10231 + }, + { + "epoch": 8.172523961661343, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 10232 + }, + { + "epoch": 8.17332268370607, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 10233 + }, + { + "epoch": 8.1741214057508, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 10234 + }, + { + "epoch": 8.174920127795527, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 10235 + }, + { + "epoch": 8.175718849840255, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 10236 + }, + { + "epoch": 8.176517571884984, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 10237 + }, + { + "epoch": 8.177316293929712, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 10238 + }, + { + "epoch": 8.178115015974441, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10239 + }, + { + "epoch": 8.178913738019169, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 10240 + }, + { + "epoch": 8.179712460063898, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 10241 + }, + { + "epoch": 8.180511182108626, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 10242 + }, + { + "epoch": 8.181309904153355, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 10243 + }, + { + "epoch": 8.182108626198083, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 10244 + }, + { + "epoch": 8.18290734824281, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 10245 + }, + { + "epoch": 8.18370607028754, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 10246 + }, + { + "epoch": 8.184504792332268, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1163, + "step": 10247 + }, + { + "epoch": 8.185303514376997, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 10248 + }, + { + "epoch": 8.186102236421725, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 10249 + }, + { + "epoch": 8.186900958466454, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 10250 + }, + { + "epoch": 8.187699680511182, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 10251 + }, + { + "epoch": 8.188498402555911, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 10252 + }, + { + "epoch": 8.189297124600639, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 10253 + }, + { + "epoch": 8.190095846645367, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 10254 + }, + { + "epoch": 8.190894568690096, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 10255 + }, + { + "epoch": 8.191693290734824, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 10256 + }, + { + "epoch": 8.192492012779553, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 10257 + }, + { + "epoch": 8.19329073482428, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 10258 + }, + { + "epoch": 8.19408945686901, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 10259 + }, + { + "epoch": 8.194888178913738, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10260 + }, + { + "epoch": 8.195686900958467, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 10261 + }, + { + "epoch": 8.196485623003195, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 10262 + }, + { + "epoch": 8.197284345047922, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 10263 + }, + { + "epoch": 8.198083067092652, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 10264 + }, + { + "epoch": 8.19888178913738, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 10265 + }, + { + "epoch": 8.199680511182109, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 10266 + }, + { + "epoch": 8.200479233226837, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 10267 + }, + { + "epoch": 8.201277955271566, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 10268 + }, + { + "epoch": 8.202076677316294, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 10269 + }, + { + "epoch": 8.202875399361023, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 10270 + }, + { + "epoch": 8.20367412140575, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 10271 + }, + { + "epoch": 8.204472843450478, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 10272 + }, + { + "epoch": 8.205271565495208, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 10273 + }, + { + "epoch": 8.206070287539935, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10274 + }, + { + "epoch": 8.206869009584665, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 10275 + }, + { + "epoch": 8.207667731629392, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 10276 + }, + { + "epoch": 8.208466453674122, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 10277 + }, + { + "epoch": 8.20926517571885, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 10278 + }, + { + "epoch": 8.210063897763579, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 10279 + }, + { + "epoch": 8.210862619808307, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1178, + "step": 10280 + }, + { + "epoch": 8.211661341853036, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 10281 + }, + { + "epoch": 8.212460063897764, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 10282 + }, + { + "epoch": 8.213258785942491, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 10283 + }, + { + "epoch": 8.21405750798722, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 10284 + }, + { + "epoch": 8.214856230031948, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 10285 + }, + { + "epoch": 8.215654952076678, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10286 + }, + { + "epoch": 8.216453674121405, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10287 + }, + { + "epoch": 8.217252396166135, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10288 + }, + { + "epoch": 8.218051118210862, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10289 + }, + { + "epoch": 8.218849840255592, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 10290 + }, + { + "epoch": 8.21964856230032, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 10291 + }, + { + "epoch": 8.220447284345047, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1176, + "step": 10292 + }, + { + "epoch": 8.221246006389777, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 10293 + }, + { + "epoch": 8.222044728434504, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 10294 + }, + { + "epoch": 8.222843450479234, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 10295 + }, + { + "epoch": 8.223642172523961, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 10296 + }, + { + "epoch": 8.22444089456869, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 10297 + }, + { + "epoch": 8.225239616613418, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 10298 + }, + { + "epoch": 8.226038338658148, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10299 + }, + { + "epoch": 8.226837060702875, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 10300 + }, + { + "epoch": 8.227635782747603, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 10301 + }, + { + "epoch": 8.228434504792332, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10302 + }, + { + "epoch": 8.22923322683706, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 10303 + }, + { + "epoch": 8.23003194888179, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10304 + }, + { + "epoch": 8.230830670926517, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 10305 + }, + { + "epoch": 8.231629392971247, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 10306 + }, + { + "epoch": 8.232428115015974, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 10307 + }, + { + "epoch": 8.233226837060704, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 10308 + }, + { + "epoch": 8.234025559105431, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 10309 + }, + { + "epoch": 8.23482428115016, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 10310 + }, + { + "epoch": 8.235623003194888, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 10311 + }, + { + "epoch": 8.236421725239616, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 10312 + }, + { + "epoch": 8.237220447284345, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 10313 + }, + { + "epoch": 8.238019169329073, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 10314 + }, + { + "epoch": 8.238817891373802, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 10315 + }, + { + "epoch": 8.23961661341853, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 10316 + }, + { + "epoch": 8.24041533546326, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 10317 + }, + { + "epoch": 8.241214057507987, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 10318 + }, + { + "epoch": 8.242012779552716, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10319 + }, + { + "epoch": 8.242811501597444, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 10320 + }, + { + "epoch": 8.243610223642172, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 10321 + }, + { + "epoch": 8.244408945686901, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10322 + }, + { + "epoch": 8.245207667731629, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 10323 + }, + { + "epoch": 8.246006389776358, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 10324 + }, + { + "epoch": 8.246805111821086, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 10325 + }, + { + "epoch": 8.247603833865815, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 10326 + }, + { + "epoch": 8.248402555910543, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 10327 + }, + { + "epoch": 8.249201277955272, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 10328 + }, + { + "epoch": 8.25, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 10329 + }, + { + "epoch": 8.250798722044728, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 10330 + }, + { + "epoch": 8.251597444089457, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 10331 + }, + { + "epoch": 8.252396166134185, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 10332 + }, + { + "epoch": 8.253194888178914, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 10333 + }, + { + "epoch": 8.253993610223642, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 10334 + }, + { + "epoch": 8.254792332268371, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 10335 + }, + { + "epoch": 8.255591054313099, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1196, + "step": 10336 + }, + { + "epoch": 8.256389776357828, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 10337 + }, + { + "epoch": 8.257188498402556, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 10338 + }, + { + "epoch": 8.257987220447284, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 10339 + }, + { + "epoch": 8.258785942492013, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10340 + }, + { + "epoch": 8.25958466453674, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 10341 + }, + { + "epoch": 8.26038338658147, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 10342 + }, + { + "epoch": 8.261182108626198, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 10343 + }, + { + "epoch": 8.261980830670927, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 10344 + }, + { + "epoch": 8.262779552715655, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 10345 + }, + { + "epoch": 8.263578274760384, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 10346 + }, + { + "epoch": 8.264376996805112, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10347 + }, + { + "epoch": 8.26517571884984, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 10348 + }, + { + "epoch": 8.265974440894569, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 10349 + }, + { + "epoch": 8.266773162939296, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 10350 + }, + { + "epoch": 8.267571884984026, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 10351 + }, + { + "epoch": 8.268370607028753, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 10352 + }, + { + "epoch": 8.269169329073483, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 10353 + }, + { + "epoch": 8.26996805111821, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 10354 + }, + { + "epoch": 8.27076677316294, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 10355 + }, + { + "epoch": 8.271565495207668, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 10356 + }, + { + "epoch": 8.272364217252397, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 10357 + }, + { + "epoch": 8.273162939297125, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 10358 + }, + { + "epoch": 8.273961661341852, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10359 + }, + { + "epoch": 8.274760383386582, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 10360 + }, + { + "epoch": 8.27555910543131, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 10361 + }, + { + "epoch": 8.276357827476039, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 10362 + }, + { + "epoch": 8.277156549520766, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 10363 + }, + { + "epoch": 8.277955271565496, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 10364 + }, + { + "epoch": 8.278753993610223, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 10365 + }, + { + "epoch": 8.279552715654953, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 10366 + }, + { + "epoch": 8.28035143769968, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10367 + }, + { + "epoch": 8.281150159744408, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 10368 + }, + { + "epoch": 8.281948881789138, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 10369 + }, + { + "epoch": 8.282747603833865, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 10370 + }, + { + "epoch": 8.283546325878595, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 10371 + }, + { + "epoch": 8.284345047923322, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 10372 + }, + { + "epoch": 8.285143769968052, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 10373 + }, + { + "epoch": 8.28594249201278, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 10374 + }, + { + "epoch": 8.286741214057509, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 10375 + }, + { + "epoch": 8.287539936102236, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 10376 + }, + { + "epoch": 8.288338658146964, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 10377 + }, + { + "epoch": 8.289137380191693, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 10378 + }, + { + "epoch": 8.289936102236421, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 10379 + }, + { + "epoch": 8.29073482428115, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 10380 + }, + { + "epoch": 8.291533546325878, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 10381 + }, + { + "epoch": 8.292332268370608, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1156, + "step": 10382 + }, + { + "epoch": 8.293130990415335, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 10383 + }, + { + "epoch": 8.293929712460065, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 10384 + }, + { + "epoch": 8.294728434504792, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 10385 + }, + { + "epoch": 8.295527156549522, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 10386 + }, + { + "epoch": 8.29632587859425, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 10387 + }, + { + "epoch": 8.297124600638977, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 10388 + }, + { + "epoch": 8.297923322683706, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1151, + "step": 10389 + }, + { + "epoch": 8.298722044728434, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 10390 + }, + { + "epoch": 8.299520766773163, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 10391 + }, + { + "epoch": 8.300319488817891, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 10392 + }, + { + "epoch": 8.30111821086262, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 10393 + }, + { + "epoch": 8.301916932907348, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 10394 + }, + { + "epoch": 8.302715654952078, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 10395 + }, + { + "epoch": 8.303514376996805, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 10396 + }, + { + "epoch": 8.304313099041533, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 10397 + }, + { + "epoch": 8.305111821086262, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10398 + }, + { + "epoch": 8.30591054313099, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 10399 + }, + { + "epoch": 8.30670926517572, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 10400 + }, + { + "epoch": 8.307507987220447, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 10401 + }, + { + "epoch": 8.308306709265176, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 10402 + }, + { + "epoch": 8.309105431309904, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 10403 + }, + { + "epoch": 8.309904153354633, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 10404 + }, + { + "epoch": 8.310702875399361, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 10405 + }, + { + "epoch": 8.311501597444089, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 10406 + }, + { + "epoch": 8.312300319488818, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 10407 + }, + { + "epoch": 8.313099041533546, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 10408 + }, + { + "epoch": 8.313897763578275, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 10409 + }, + { + "epoch": 8.314696485623003, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 10410 + }, + { + "epoch": 8.315495207667732, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 10411 + }, + { + "epoch": 8.31629392971246, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10412 + }, + { + "epoch": 8.31709265175719, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 10413 + }, + { + "epoch": 8.317891373801917, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10414 + }, + { + "epoch": 8.318690095846645, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 10415 + }, + { + "epoch": 8.319488817891374, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10416 + }, + { + "epoch": 8.320287539936102, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 10417 + }, + { + "epoch": 8.321086261980831, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1173, + "step": 10418 + }, + { + "epoch": 8.321884984025559, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 10419 + }, + { + "epoch": 8.322683706070288, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 10420 + }, + { + "epoch": 8.323482428115016, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 10421 + }, + { + "epoch": 8.324281150159745, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 10422 + }, + { + "epoch": 8.325079872204473, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 10423 + }, + { + "epoch": 8.3258785942492, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 10424 + }, + { + "epoch": 8.32667731629393, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 10425 + }, + { + "epoch": 8.327476038338657, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 10426 + }, + { + "epoch": 8.328274760383387, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 10427 + }, + { + "epoch": 8.329073482428115, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 10428 + }, + { + "epoch": 8.329872204472844, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 10429 + }, + { + "epoch": 8.330670926517572, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 10430 + }, + { + "epoch": 8.331469648562301, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 10431 + }, + { + "epoch": 8.332268370607029, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 10432 + }, + { + "epoch": 8.333067092651758, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 10433 + }, + { + "epoch": 8.333865814696486, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 10434 + }, + { + "epoch": 8.334664536741213, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 10435 + }, + { + "epoch": 8.335463258785943, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 10436 + }, + { + "epoch": 8.33626198083067, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1146, + "step": 10437 + }, + { + "epoch": 8.3370607028754, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 10438 + }, + { + "epoch": 8.337859424920127, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 10439 + }, + { + "epoch": 8.338658146964857, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 10440 + }, + { + "epoch": 8.339456869009584, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 10441 + }, + { + "epoch": 8.340255591054314, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 10442 + }, + { + "epoch": 8.341054313099042, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 10443 + }, + { + "epoch": 8.34185303514377, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 10444 + }, + { + "epoch": 8.342651757188499, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 10445 + }, + { + "epoch": 8.343450479233226, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 10446 + }, + { + "epoch": 8.344249201277956, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 10447 + }, + { + "epoch": 8.345047923322683, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 10448 + }, + { + "epoch": 8.345846645367413, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10449 + }, + { + "epoch": 8.34664536741214, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 10450 + }, + { + "epoch": 8.34744408945687, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 10451 + }, + { + "epoch": 8.348242811501597, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 10452 + }, + { + "epoch": 8.349041533546325, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 10453 + }, + { + "epoch": 8.349840255591054, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 10454 + }, + { + "epoch": 8.350638977635782, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.117, + "step": 10455 + }, + { + "epoch": 8.351437699680512, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 10456 + }, + { + "epoch": 8.35223642172524, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 10457 + }, + { + "epoch": 8.353035143769969, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 10458 + }, + { + "epoch": 8.353833865814696, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 10459 + }, + { + "epoch": 8.354632587859426, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 10460 + }, + { + "epoch": 8.355431309904153, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 10461 + }, + { + "epoch": 8.356230031948883, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 10462 + }, + { + "epoch": 8.35702875399361, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 10463 + }, + { + "epoch": 8.357827476038338, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 10464 + }, + { + "epoch": 8.358626198083067, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 10465 + }, + { + "epoch": 8.359424920127795, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 10466 + }, + { + "epoch": 8.360223642172524, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 10467 + }, + { + "epoch": 8.361022364217252, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 10468 + }, + { + "epoch": 8.361821086261982, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 10469 + }, + { + "epoch": 8.36261980830671, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 10470 + }, + { + "epoch": 8.363418530351439, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 10471 + }, + { + "epoch": 8.364217252396166, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10472 + }, + { + "epoch": 8.365015974440894, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 10473 + }, + { + "epoch": 8.365814696485623, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 10474 + }, + { + "epoch": 8.366613418530351, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 10475 + }, + { + "epoch": 8.36741214057508, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 10476 + }, + { + "epoch": 8.368210862619808, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 10477 + }, + { + "epoch": 8.369009584664537, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10478 + }, + { + "epoch": 8.369808306709265, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 10479 + }, + { + "epoch": 8.370607028753994, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 10480 + }, + { + "epoch": 8.371405750798722, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 10481 + }, + { + "epoch": 8.37220447284345, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 10482 + }, + { + "epoch": 8.37300319488818, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 10483 + }, + { + "epoch": 8.373801916932907, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10484 + }, + { + "epoch": 8.374600638977636, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10485 + }, + { + "epoch": 8.375399361022364, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 10486 + }, + { + "epoch": 8.376198083067093, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 10487 + }, + { + "epoch": 8.37699680511182, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 10488 + }, + { + "epoch": 8.37779552715655, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 10489 + }, + { + "epoch": 8.378594249201278, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 10490 + }, + { + "epoch": 8.379392971246006, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 10491 + }, + { + "epoch": 8.380191693290735, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 10492 + }, + { + "epoch": 8.380990415335463, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 10493 + }, + { + "epoch": 8.381789137380192, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 10494 + }, + { + "epoch": 8.38258785942492, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 10495 + }, + { + "epoch": 8.383386581469649, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 10496 + }, + { + "epoch": 8.384185303514377, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 10497 + }, + { + "epoch": 8.384984025559106, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 10498 + }, + { + "epoch": 8.385782747603834, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 10499 + }, + { + "epoch": 8.386581469648561, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 10500 + }, + { + "epoch": 8.38738019169329, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 10501 + }, + { + "epoch": 8.388178913738018, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 10502 + }, + { + "epoch": 8.388977635782748, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 10503 + }, + { + "epoch": 8.389776357827476, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10504 + }, + { + "epoch": 8.390575079872205, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 10505 + }, + { + "epoch": 8.391373801916933, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 10506 + }, + { + "epoch": 8.392172523961662, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 10507 + }, + { + "epoch": 8.39297124600639, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 10508 + }, + { + "epoch": 8.393769968051117, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 10509 + }, + { + "epoch": 8.394568690095847, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 10510 + }, + { + "epoch": 8.395367412140574, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 10511 + }, + { + "epoch": 8.396166134185304, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 10512 + }, + { + "epoch": 8.396964856230031, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 10513 + }, + { + "epoch": 8.39776357827476, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 10514 + }, + { + "epoch": 8.398562300319488, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 10515 + }, + { + "epoch": 8.399361022364218, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10516 + }, + { + "epoch": 8.400159744408946, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 10517 + }, + { + "epoch": 8.400958466453675, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 10518 + }, + { + "epoch": 8.401757188498403, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 10519 + }, + { + "epoch": 8.40255591054313, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 10520 + }, + { + "epoch": 8.40335463258786, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10521 + }, + { + "epoch": 8.404153354632587, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 10522 + }, + { + "epoch": 8.404952076677317, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 10523 + }, + { + "epoch": 8.405750798722044, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 10524 + }, + { + "epoch": 8.406549520766774, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 10525 + }, + { + "epoch": 8.407348242811501, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 10526 + }, + { + "epoch": 8.40814696485623, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 10527 + }, + { + "epoch": 8.408945686900958, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 10528 + }, + { + "epoch": 8.409744408945686, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 10529 + }, + { + "epoch": 8.410543130990416, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 10530 + }, + { + "epoch": 8.411341853035143, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 10531 + }, + { + "epoch": 8.412140575079873, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 10532 + }, + { + "epoch": 8.4129392971246, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 10533 + }, + { + "epoch": 8.41373801916933, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 10534 + }, + { + "epoch": 8.414536741214057, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 10535 + }, + { + "epoch": 8.415335463258787, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 10536 + }, + { + "epoch": 8.416134185303514, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 10537 + }, + { + "epoch": 8.416932907348242, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 10538 + }, + { + "epoch": 8.417731629392971, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 10539 + }, + { + "epoch": 8.418530351437699, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 10540 + }, + { + "epoch": 8.419329073482428, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10541 + }, + { + "epoch": 8.420127795527156, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10542 + }, + { + "epoch": 8.420926517571885, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 10543 + }, + { + "epoch": 8.421725239616613, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 10544 + }, + { + "epoch": 8.422523961661343, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 10545 + }, + { + "epoch": 8.42332268370607, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10546 + }, + { + "epoch": 8.4241214057508, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 10547 + }, + { + "epoch": 8.424920127795527, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 10548 + }, + { + "epoch": 8.425718849840255, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 10549 + }, + { + "epoch": 8.426517571884984, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 10550 + }, + { + "epoch": 8.427316293929712, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 10551 + }, + { + "epoch": 8.428115015974441, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 10552 + }, + { + "epoch": 8.428913738019169, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 10553 + }, + { + "epoch": 8.429712460063898, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 10554 + }, + { + "epoch": 8.430511182108626, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 10555 + }, + { + "epoch": 8.431309904153355, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 10556 + }, + { + "epoch": 8.432108626198083, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 10557 + }, + { + "epoch": 8.43290734824281, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 10558 + }, + { + "epoch": 8.43370607028754, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 10559 + }, + { + "epoch": 8.434504792332268, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 10560 + }, + { + "epoch": 8.435303514376997, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 10561 + }, + { + "epoch": 8.436102236421725, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 10562 + }, + { + "epoch": 8.436900958466454, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 10563 + }, + { + "epoch": 8.437699680511182, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 10564 + }, + { + "epoch": 8.438498402555911, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 10565 + }, + { + "epoch": 8.439297124600639, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 10566 + }, + { + "epoch": 8.440095846645367, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 10567 + }, + { + "epoch": 8.440894568690096, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 10568 + }, + { + "epoch": 8.441693290734824, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 10569 + }, + { + "epoch": 8.442492012779553, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 10570 + }, + { + "epoch": 8.44329073482428, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 10571 + }, + { + "epoch": 8.44408945686901, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 10572 + }, + { + "epoch": 8.444888178913738, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 10573 + }, + { + "epoch": 8.445686900958467, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 10574 + }, + { + "epoch": 8.446485623003195, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 10575 + }, + { + "epoch": 8.447284345047922, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 10576 + }, + { + "epoch": 8.448083067092652, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 10577 + }, + { + "epoch": 8.44888178913738, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 10578 + }, + { + "epoch": 8.449680511182109, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 10579 + }, + { + "epoch": 8.450479233226837, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 10580 + }, + { + "epoch": 8.451277955271566, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 10581 + }, + { + "epoch": 8.452076677316294, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 10582 + }, + { + "epoch": 8.452875399361023, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 10583 + }, + { + "epoch": 8.45367412140575, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 10584 + }, + { + "epoch": 8.454472843450478, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 10585 + }, + { + "epoch": 8.455271565495208, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 10586 + }, + { + "epoch": 8.456070287539935, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 10587 + }, + { + "epoch": 8.456869009584665, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1142, + "step": 10588 + }, + { + "epoch": 8.457667731629392, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 10589 + }, + { + "epoch": 8.458466453674122, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 10590 + }, + { + "epoch": 8.45926517571885, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 10591 + }, + { + "epoch": 8.460063897763579, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 10592 + }, + { + "epoch": 8.460862619808307, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 10593 + }, + { + "epoch": 8.461661341853036, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 10594 + }, + { + "epoch": 8.462460063897764, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 10595 + }, + { + "epoch": 8.463258785942491, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 10596 + }, + { + "epoch": 8.46405750798722, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10597 + }, + { + "epoch": 8.464856230031948, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 10598 + }, + { + "epoch": 8.465654952076678, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 10599 + }, + { + "epoch": 8.466453674121405, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 10600 + }, + { + "epoch": 8.467252396166135, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 10601 + }, + { + "epoch": 8.468051118210862, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 10602 + }, + { + "epoch": 8.468849840255592, + "grad_norm": 0.365234375, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 10603 + }, + { + "epoch": 8.46964856230032, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10604 + }, + { + "epoch": 8.470447284345047, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 10605 + }, + { + "epoch": 8.471246006389777, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 10606 + }, + { + "epoch": 8.472044728434504, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 10607 + }, + { + "epoch": 8.472843450479234, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 10608 + }, + { + "epoch": 8.473642172523961, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 10609 + }, + { + "epoch": 8.47444089456869, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 10610 + }, + { + "epoch": 8.475239616613418, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10611 + }, + { + "epoch": 8.476038338658148, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 10612 + }, + { + "epoch": 8.476837060702875, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 10613 + }, + { + "epoch": 8.477635782747603, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 10614 + }, + { + "epoch": 8.478434504792332, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 10615 + }, + { + "epoch": 8.47923322683706, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 10616 + }, + { + "epoch": 8.48003194888179, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 10617 + }, + { + "epoch": 8.480830670926517, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1162, + "step": 10618 + }, + { + "epoch": 8.481629392971247, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 10619 + }, + { + "epoch": 8.482428115015974, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 10620 + }, + { + "epoch": 8.483226837060704, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 10621 + }, + { + "epoch": 8.484025559105431, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 10622 + }, + { + "epoch": 8.48482428115016, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 10623 + }, + { + "epoch": 8.485623003194888, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 10624 + }, + { + "epoch": 8.486421725239616, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 10625 + }, + { + "epoch": 8.487220447284345, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 10626 + }, + { + "epoch": 8.488019169329073, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 10627 + }, + { + "epoch": 8.488817891373802, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 10628 + }, + { + "epoch": 8.48961661341853, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 10629 + }, + { + "epoch": 8.49041533546326, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 10630 + }, + { + "epoch": 8.491214057507987, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 10631 + }, + { + "epoch": 8.492012779552716, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 10632 + }, + { + "epoch": 8.492811501597444, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 10633 + }, + { + "epoch": 8.493610223642172, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 10634 + }, + { + "epoch": 8.494408945686901, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10635 + }, + { + "epoch": 8.495207667731629, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 10636 + }, + { + "epoch": 8.496006389776358, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 10637 + }, + { + "epoch": 8.496805111821086, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 10638 + }, + { + "epoch": 8.497603833865815, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 10639 + }, + { + "epoch": 8.498402555910543, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 10640 + }, + { + "epoch": 8.499201277955272, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10641 + }, + { + "epoch": 8.5, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 10642 + }, + { + "epoch": 8.500798722044728, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 10643 + }, + { + "epoch": 8.501597444089457, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 10644 + }, + { + "epoch": 8.502396166134185, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 10645 + }, + { + "epoch": 8.503194888178914, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 10646 + }, + { + "epoch": 8.503993610223642, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 10647 + }, + { + "epoch": 8.504792332268371, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 10648 + }, + { + "epoch": 8.505591054313099, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 10649 + }, + { + "epoch": 8.506389776357828, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 10650 + }, + { + "epoch": 8.507188498402556, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 10651 + }, + { + "epoch": 8.507987220447284, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 10652 + }, + { + "epoch": 8.508785942492013, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 10653 + }, + { + "epoch": 8.50958466453674, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10654 + }, + { + "epoch": 8.51038338658147, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 10655 + }, + { + "epoch": 8.511182108626198, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 10656 + }, + { + "epoch": 8.511980830670927, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 10657 + }, + { + "epoch": 8.512779552715655, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 10658 + }, + { + "epoch": 8.513578274760384, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10659 + }, + { + "epoch": 8.514376996805112, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 10660 + }, + { + "epoch": 8.51517571884984, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 10661 + }, + { + "epoch": 8.515974440894569, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 10662 + }, + { + "epoch": 8.516773162939296, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 10663 + }, + { + "epoch": 8.517571884984026, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 10664 + }, + { + "epoch": 8.518370607028753, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 10665 + }, + { + "epoch": 8.519169329073483, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 10666 + }, + { + "epoch": 8.51996805111821, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 10667 + }, + { + "epoch": 8.52076677316294, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 10668 + }, + { + "epoch": 8.521565495207668, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 10669 + }, + { + "epoch": 8.522364217252395, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 10670 + }, + { + "epoch": 8.523162939297125, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 10671 + }, + { + "epoch": 8.523961661341852, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 10672 + }, + { + "epoch": 8.524760383386582, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 10673 + }, + { + "epoch": 8.52555910543131, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 10674 + }, + { + "epoch": 8.526357827476039, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 10675 + }, + { + "epoch": 8.527156549520766, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 10676 + }, + { + "epoch": 8.527955271565496, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 10677 + }, + { + "epoch": 8.528753993610223, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 10678 + }, + { + "epoch": 8.529552715654953, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 10679 + }, + { + "epoch": 8.53035143769968, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 10680 + }, + { + "epoch": 8.531150159744408, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 10681 + }, + { + "epoch": 8.531948881789138, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 10682 + }, + { + "epoch": 8.532747603833865, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 10683 + }, + { + "epoch": 8.533546325878595, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 10684 + }, + { + "epoch": 8.534345047923322, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 10685 + }, + { + "epoch": 8.535143769968052, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 10686 + }, + { + "epoch": 8.53594249201278, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 10687 + }, + { + "epoch": 8.536741214057509, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 10688 + }, + { + "epoch": 8.537539936102236, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 10689 + }, + { + "epoch": 8.538338658146966, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 10690 + }, + { + "epoch": 8.539137380191693, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 10691 + }, + { + "epoch": 8.539936102236421, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 10692 + }, + { + "epoch": 8.54073482428115, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 10693 + }, + { + "epoch": 8.541533546325878, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 10694 + }, + { + "epoch": 8.542332268370608, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 10695 + }, + { + "epoch": 8.543130990415335, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 10696 + }, + { + "epoch": 8.543929712460065, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 10697 + }, + { + "epoch": 8.544728434504792, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 10698 + }, + { + "epoch": 8.545527156549522, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 10699 + }, + { + "epoch": 8.54632587859425, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 10700 + }, + { + "epoch": 8.547124600638977, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 10701 + }, + { + "epoch": 8.547923322683706, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 10702 + }, + { + "epoch": 8.548722044728434, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 10703 + }, + { + "epoch": 8.549520766773163, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 10704 + }, + { + "epoch": 8.550319488817891, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 10705 + }, + { + "epoch": 8.55111821086262, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 10706 + }, + { + "epoch": 8.551916932907348, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 10707 + }, + { + "epoch": 8.552715654952078, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 10708 + }, + { + "epoch": 8.553514376996805, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 10709 + }, + { + "epoch": 8.554313099041533, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 10710 + }, + { + "epoch": 8.555111821086262, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 10711 + }, + { + "epoch": 8.55591054313099, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 10712 + }, + { + "epoch": 8.55670926517572, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 10713 + }, + { + "epoch": 8.557507987220447, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 10714 + }, + { + "epoch": 8.558306709265176, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 10715 + }, + { + "epoch": 8.559105431309904, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 10716 + }, + { + "epoch": 8.559904153354633, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 10717 + }, + { + "epoch": 8.560702875399361, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 10718 + }, + { + "epoch": 8.561501597444089, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 10719 + }, + { + "epoch": 8.562300319488818, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 10720 + }, + { + "epoch": 8.563099041533546, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10721 + }, + { + "epoch": 8.563897763578275, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 10722 + }, + { + "epoch": 8.564696485623003, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 10723 + }, + { + "epoch": 8.565495207667732, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10724 + }, + { + "epoch": 8.56629392971246, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10725 + }, + { + "epoch": 8.56709265175719, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 10726 + }, + { + "epoch": 8.567891373801917, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 10727 + }, + { + "epoch": 8.568690095846645, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 10728 + }, + { + "epoch": 8.569488817891374, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 10729 + }, + { + "epoch": 8.570287539936102, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 10730 + }, + { + "epoch": 8.571086261980831, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 10731 + }, + { + "epoch": 8.571884984025559, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 10732 + }, + { + "epoch": 8.572683706070288, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 10733 + }, + { + "epoch": 8.573482428115016, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 10734 + }, + { + "epoch": 8.574281150159745, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10735 + }, + { + "epoch": 8.575079872204473, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 10736 + }, + { + "epoch": 8.5758785942492, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 10737 + }, + { + "epoch": 8.57667731629393, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 10738 + }, + { + "epoch": 8.577476038338657, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 10739 + }, + { + "epoch": 8.578274760383387, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 10740 + }, + { + "epoch": 8.579073482428115, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 10741 + }, + { + "epoch": 8.579872204472844, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10742 + }, + { + "epoch": 8.580670926517572, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 10743 + }, + { + "epoch": 8.581469648562301, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 10744 + }, + { + "epoch": 8.582268370607029, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 10745 + }, + { + "epoch": 8.583067092651756, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 10746 + }, + { + "epoch": 8.583865814696486, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 10747 + }, + { + "epoch": 8.584664536741213, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 10748 + }, + { + "epoch": 8.585463258785943, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 10749 + }, + { + "epoch": 8.58626198083067, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 10750 + }, + { + "epoch": 8.5870607028754, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 10751 + }, + { + "epoch": 8.587859424920127, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10752 + }, + { + "epoch": 8.588658146964857, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 10753 + }, + { + "epoch": 8.589456869009584, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 10754 + }, + { + "epoch": 8.590255591054314, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 10755 + }, + { + "epoch": 8.591054313099042, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10756 + }, + { + "epoch": 8.59185303514377, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 10757 + }, + { + "epoch": 8.592651757188499, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 10758 + }, + { + "epoch": 8.593450479233226, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 10759 + }, + { + "epoch": 8.594249201277956, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10760 + }, + { + "epoch": 8.595047923322683, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 10761 + }, + { + "epoch": 8.595846645367413, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 10762 + }, + { + "epoch": 8.59664536741214, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 10763 + }, + { + "epoch": 8.59744408945687, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 10764 + }, + { + "epoch": 8.598242811501597, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 10765 + }, + { + "epoch": 8.599041533546325, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 10766 + }, + { + "epoch": 8.599840255591054, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 10767 + }, + { + "epoch": 8.600638977635782, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 10768 + }, + { + "epoch": 8.601437699680512, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 10769 + }, + { + "epoch": 8.60223642172524, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 10770 + }, + { + "epoch": 8.603035143769969, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 10771 + }, + { + "epoch": 8.603833865814696, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 10772 + }, + { + "epoch": 8.604632587859426, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 10773 + }, + { + "epoch": 8.605431309904153, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 10774 + }, + { + "epoch": 8.606230031948883, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 10775 + }, + { + "epoch": 8.60702875399361, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 10776 + }, + { + "epoch": 8.607827476038338, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 10777 + }, + { + "epoch": 8.608626198083067, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 10778 + }, + { + "epoch": 8.609424920127795, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 10779 + }, + { + "epoch": 8.610223642172524, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 10780 + }, + { + "epoch": 8.611022364217252, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 10781 + }, + { + "epoch": 8.611821086261982, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 10782 + }, + { + "epoch": 8.61261980830671, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 10783 + }, + { + "epoch": 8.613418530351439, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 10784 + }, + { + "epoch": 8.614217252396166, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 10785 + }, + { + "epoch": 8.615015974440894, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 10786 + }, + { + "epoch": 8.615814696485623, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 10787 + }, + { + "epoch": 8.616613418530351, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 10788 + }, + { + "epoch": 8.61741214057508, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 10789 + }, + { + "epoch": 8.618210862619808, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 10790 + }, + { + "epoch": 8.619009584664537, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 10791 + }, + { + "epoch": 8.619808306709265, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 10792 + }, + { + "epoch": 8.620607028753994, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1199, + "step": 10793 + }, + { + "epoch": 8.621405750798722, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 10794 + }, + { + "epoch": 8.62220447284345, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 10795 + }, + { + "epoch": 8.62300319488818, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 10796 + }, + { + "epoch": 8.623801916932907, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 10797 + }, + { + "epoch": 8.624600638977636, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10798 + }, + { + "epoch": 8.625399361022364, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 10799 + }, + { + "epoch": 8.626198083067093, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 10800 + }, + { + "epoch": 8.62699680511182, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 10801 + }, + { + "epoch": 8.62779552715655, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 10802 + }, + { + "epoch": 8.628594249201278, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1138, + "step": 10803 + }, + { + "epoch": 8.629392971246006, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10804 + }, + { + "epoch": 8.630191693290735, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 10805 + }, + { + "epoch": 8.630990415335463, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10806 + }, + { + "epoch": 8.631789137380192, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10807 + }, + { + "epoch": 8.63258785942492, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 10808 + }, + { + "epoch": 8.633386581469649, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 10809 + }, + { + "epoch": 8.634185303514377, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 10810 + }, + { + "epoch": 8.634984025559106, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 10811 + }, + { + "epoch": 8.635782747603834, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10812 + }, + { + "epoch": 8.636581469648561, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 10813 + }, + { + "epoch": 8.63738019169329, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 10814 + }, + { + "epoch": 8.638178913738018, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 10815 + }, + { + "epoch": 8.638977635782748, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 10816 + }, + { + "epoch": 8.639776357827476, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 10817 + }, + { + "epoch": 8.640575079872205, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 10818 + }, + { + "epoch": 8.641373801916933, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 10819 + }, + { + "epoch": 8.642172523961662, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10820 + }, + { + "epoch": 8.64297124600639, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 10821 + }, + { + "epoch": 8.643769968051117, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 10822 + }, + { + "epoch": 8.644568690095847, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 10823 + }, + { + "epoch": 8.645367412140574, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 10824 + }, + { + "epoch": 8.646166134185304, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 10825 + }, + { + "epoch": 8.646964856230031, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 10826 + }, + { + "epoch": 8.64776357827476, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10827 + }, + { + "epoch": 8.648562300319488, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 10828 + }, + { + "epoch": 8.649361022364218, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10829 + }, + { + "epoch": 8.650159744408946, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 10830 + }, + { + "epoch": 8.650958466453675, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 10831 + }, + { + "epoch": 8.651757188498403, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 10832 + }, + { + "epoch": 8.65255591054313, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 10833 + }, + { + "epoch": 8.65335463258786, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 10834 + }, + { + "epoch": 8.654153354632587, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 10835 + }, + { + "epoch": 8.654952076677317, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 10836 + }, + { + "epoch": 8.655750798722044, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 10837 + }, + { + "epoch": 8.656549520766774, + "grad_norm": 0.09814453125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 10838 + }, + { + "epoch": 8.657348242811501, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 10839 + }, + { + "epoch": 8.65814696485623, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 10840 + }, + { + "epoch": 8.658945686900958, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 10841 + }, + { + "epoch": 8.659744408945686, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 10842 + }, + { + "epoch": 8.660543130990416, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 10843 + }, + { + "epoch": 8.661341853035143, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 10844 + }, + { + "epoch": 8.662140575079873, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 10845 + }, + { + "epoch": 8.6629392971246, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 10846 + }, + { + "epoch": 8.66373801916933, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 10847 + }, + { + "epoch": 8.664536741214057, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 10848 + }, + { + "epoch": 8.665335463258787, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 10849 + }, + { + "epoch": 8.666134185303514, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 10850 + }, + { + "epoch": 8.666932907348244, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 10851 + }, + { + "epoch": 8.667731629392971, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 10852 + }, + { + "epoch": 8.668530351437699, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 10853 + }, + { + "epoch": 8.669329073482428, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10854 + }, + { + "epoch": 8.670127795527156, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 10855 + }, + { + "epoch": 8.670926517571885, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 10856 + }, + { + "epoch": 8.671725239616613, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 10857 + }, + { + "epoch": 8.672523961661343, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 10858 + }, + { + "epoch": 8.67332268370607, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 10859 + }, + { + "epoch": 8.6741214057508, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 10860 + }, + { + "epoch": 8.674920127795527, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 10861 + }, + { + "epoch": 8.675718849840255, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 10862 + }, + { + "epoch": 8.676517571884984, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 10863 + }, + { + "epoch": 8.677316293929712, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 10864 + }, + { + "epoch": 8.678115015974441, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 10865 + }, + { + "epoch": 8.678913738019169, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 10866 + }, + { + "epoch": 8.679712460063898, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 10867 + }, + { + "epoch": 8.680511182108626, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 10868 + }, + { + "epoch": 8.681309904153355, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 10869 + }, + { + "epoch": 8.682108626198083, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 10870 + }, + { + "epoch": 8.68290734824281, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 10871 + }, + { + "epoch": 8.68370607028754, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 10872 + }, + { + "epoch": 8.684504792332268, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 10873 + }, + { + "epoch": 8.685303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 10874 + }, + { + "epoch": 8.686102236421725, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 10875 + }, + { + "epoch": 8.686900958466454, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 10876 + }, + { + "epoch": 8.687699680511182, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 10877 + }, + { + "epoch": 8.688498402555911, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 10878 + }, + { + "epoch": 8.689297124600639, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 10879 + }, + { + "epoch": 8.690095846645367, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 10880 + }, + { + "epoch": 8.690894568690096, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 10881 + }, + { + "epoch": 8.691693290734824, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 10882 + }, + { + "epoch": 8.692492012779553, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 10883 + }, + { + "epoch": 8.69329073482428, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 10884 + }, + { + "epoch": 8.69408945686901, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 10885 + }, + { + "epoch": 8.694888178913738, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 10886 + }, + { + "epoch": 8.695686900958467, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 10887 + }, + { + "epoch": 8.696485623003195, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 10888 + }, + { + "epoch": 8.697284345047922, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 10889 + }, + { + "epoch": 8.698083067092652, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 10890 + }, + { + "epoch": 8.69888178913738, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 10891 + }, + { + "epoch": 8.699680511182109, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 10892 + }, + { + "epoch": 8.700479233226837, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 10893 + }, + { + "epoch": 8.701277955271566, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 10894 + }, + { + "epoch": 8.702076677316294, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 10895 + }, + { + "epoch": 8.702875399361023, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 10896 + }, + { + "epoch": 8.70367412140575, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 10897 + }, + { + "epoch": 8.704472843450478, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 10898 + }, + { + "epoch": 8.705271565495208, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 10899 + }, + { + "epoch": 8.706070287539935, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10900 + }, + { + "epoch": 8.706869009584665, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 10901 + }, + { + "epoch": 8.707667731629392, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 10902 + }, + { + "epoch": 8.708466453674122, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 10903 + }, + { + "epoch": 8.70926517571885, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 10904 + }, + { + "epoch": 8.710063897763579, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 10905 + }, + { + "epoch": 8.710862619808307, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 10906 + }, + { + "epoch": 8.711661341853034, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 10907 + }, + { + "epoch": 8.712460063897764, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 10908 + }, + { + "epoch": 8.713258785942491, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 10909 + }, + { + "epoch": 8.71405750798722, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 10910 + }, + { + "epoch": 8.714856230031948, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 10911 + }, + { + "epoch": 8.715654952076678, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 10912 + }, + { + "epoch": 8.716453674121405, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 10913 + }, + { + "epoch": 8.717252396166135, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 10914 + }, + { + "epoch": 8.718051118210862, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 10915 + }, + { + "epoch": 8.718849840255592, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10916 + }, + { + "epoch": 8.71964856230032, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 10917 + }, + { + "epoch": 8.720447284345047, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 10918 + }, + { + "epoch": 8.721246006389777, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 10919 + }, + { + "epoch": 8.722044728434504, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 10920 + }, + { + "epoch": 8.722843450479234, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 10921 + }, + { + "epoch": 8.723642172523961, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 10922 + }, + { + "epoch": 8.72444089456869, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 10923 + }, + { + "epoch": 8.725239616613418, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 10924 + }, + { + "epoch": 8.726038338658148, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 10925 + }, + { + "epoch": 8.726837060702875, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 10926 + }, + { + "epoch": 8.727635782747605, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 10927 + }, + { + "epoch": 8.728434504792332, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 10928 + }, + { + "epoch": 8.72923322683706, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 10929 + }, + { + "epoch": 8.73003194888179, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 10930 + }, + { + "epoch": 8.730830670926517, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 10931 + }, + { + "epoch": 8.731629392971247, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 10932 + }, + { + "epoch": 8.732428115015974, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10933 + }, + { + "epoch": 8.733226837060704, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 10934 + }, + { + "epoch": 8.734025559105431, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 10935 + }, + { + "epoch": 8.73482428115016, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 10936 + }, + { + "epoch": 8.735623003194888, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 10937 + }, + { + "epoch": 8.736421725239616, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 10938 + }, + { + "epoch": 8.737220447284345, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 10939 + }, + { + "epoch": 8.738019169329073, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10940 + }, + { + "epoch": 8.738817891373802, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 10941 + }, + { + "epoch": 8.73961661341853, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 10942 + }, + { + "epoch": 8.74041533546326, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 10943 + }, + { + "epoch": 8.741214057507987, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 10944 + }, + { + "epoch": 8.742012779552716, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 10945 + }, + { + "epoch": 8.742811501597444, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 10946 + }, + { + "epoch": 8.743610223642172, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 10947 + }, + { + "epoch": 8.744408945686901, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 10948 + }, + { + "epoch": 8.745207667731629, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 10949 + }, + { + "epoch": 8.746006389776358, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10950 + }, + { + "epoch": 8.746805111821086, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 10951 + }, + { + "epoch": 8.747603833865815, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 10952 + }, + { + "epoch": 8.748402555910543, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 10953 + }, + { + "epoch": 8.749201277955272, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 10954 + }, + { + "epoch": 8.75, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 10955 + }, + { + "epoch": 8.750798722044728, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 10956 + }, + { + "epoch": 8.751597444089457, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 10957 + }, + { + "epoch": 8.752396166134185, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 10958 + }, + { + "epoch": 8.753194888178914, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 10959 + }, + { + "epoch": 8.753993610223642, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 10960 + }, + { + "epoch": 8.754792332268371, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 10961 + }, + { + "epoch": 8.755591054313099, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 10962 + }, + { + "epoch": 8.756389776357828, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 10963 + }, + { + "epoch": 8.757188498402556, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 10964 + }, + { + "epoch": 8.757987220447284, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 10965 + }, + { + "epoch": 8.758785942492013, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 10966 + }, + { + "epoch": 8.75958466453674, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 10967 + }, + { + "epoch": 8.76038338658147, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 10968 + }, + { + "epoch": 8.761182108626198, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 10969 + }, + { + "epoch": 8.761980830670927, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 10970 + }, + { + "epoch": 8.762779552715655, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 10971 + }, + { + "epoch": 8.763578274760384, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 10972 + }, + { + "epoch": 8.764376996805112, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 10973 + }, + { + "epoch": 8.76517571884984, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 10974 + }, + { + "epoch": 8.765974440894569, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 10975 + }, + { + "epoch": 8.766773162939296, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 10976 + }, + { + "epoch": 8.767571884984026, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 10977 + }, + { + "epoch": 8.768370607028753, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 10978 + }, + { + "epoch": 8.769169329073483, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 10979 + }, + { + "epoch": 8.76996805111821, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 10980 + }, + { + "epoch": 8.77076677316294, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 10981 + }, + { + "epoch": 8.771565495207668, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 10982 + }, + { + "epoch": 8.772364217252395, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 10983 + }, + { + "epoch": 8.773162939297125, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 10984 + }, + { + "epoch": 8.773961661341852, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 10985 + }, + { + "epoch": 8.774760383386582, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 10986 + }, + { + "epoch": 8.77555910543131, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 10987 + }, + { + "epoch": 8.776357827476039, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 10988 + }, + { + "epoch": 8.777156549520766, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 10989 + }, + { + "epoch": 8.777955271565496, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 10990 + }, + { + "epoch": 8.778753993610223, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 10991 + }, + { + "epoch": 8.779552715654953, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 10992 + }, + { + "epoch": 8.78035143769968, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 10993 + }, + { + "epoch": 8.781150159744408, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 10994 + }, + { + "epoch": 8.781948881789138, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 10995 + }, + { + "epoch": 8.782747603833865, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 10996 + }, + { + "epoch": 8.783546325878595, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 10997 + }, + { + "epoch": 8.784345047923322, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 10998 + }, + { + "epoch": 8.785143769968052, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 10999 + }, + { + "epoch": 8.78594249201278, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 11000 + }, + { + "epoch": 8.786741214057509, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 11001 + }, + { + "epoch": 8.787539936102236, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 11002 + }, + { + "epoch": 8.788338658146966, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 11003 + }, + { + "epoch": 8.789137380191693, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 11004 + }, + { + "epoch": 8.789936102236421, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 11005 + }, + { + "epoch": 8.79073482428115, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 11006 + }, + { + "epoch": 8.791533546325878, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 11007 + }, + { + "epoch": 8.792332268370608, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 11008 + }, + { + "epoch": 8.793130990415335, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 11009 + }, + { + "epoch": 8.793929712460065, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11010 + }, + { + "epoch": 8.794728434504792, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 11011 + }, + { + "epoch": 8.795527156549522, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 11012 + }, + { + "epoch": 8.79632587859425, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 11013 + }, + { + "epoch": 8.797124600638977, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 11014 + }, + { + "epoch": 8.797923322683706, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11015 + }, + { + "epoch": 8.798722044728434, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11016 + }, + { + "epoch": 8.799520766773163, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 11017 + }, + { + "epoch": 8.800319488817891, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 11018 + }, + { + "epoch": 8.80111821086262, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 11019 + }, + { + "epoch": 8.801916932907348, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 11020 + }, + { + "epoch": 8.802715654952078, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 11021 + }, + { + "epoch": 8.803514376996805, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 11022 + }, + { + "epoch": 8.804313099041533, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 11023 + }, + { + "epoch": 8.805111821086262, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 11024 + }, + { + "epoch": 8.80591054313099, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 11025 + }, + { + "epoch": 8.80670926517572, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 11026 + }, + { + "epoch": 8.807507987220447, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 11027 + }, + { + "epoch": 8.808306709265176, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 11028 + }, + { + "epoch": 8.809105431309904, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 11029 + }, + { + "epoch": 8.809904153354633, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 11030 + }, + { + "epoch": 8.810702875399361, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 11031 + }, + { + "epoch": 8.811501597444089, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 11032 + }, + { + "epoch": 8.812300319488818, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 11033 + }, + { + "epoch": 8.813099041533546, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 11034 + }, + { + "epoch": 8.813897763578275, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 11035 + }, + { + "epoch": 8.814696485623003, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 11036 + }, + { + "epoch": 8.815495207667732, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 11037 + }, + { + "epoch": 8.81629392971246, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 11038 + }, + { + "epoch": 8.81709265175719, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 11039 + }, + { + "epoch": 8.817891373801917, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 11040 + }, + { + "epoch": 8.818690095846645, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 11041 + }, + { + "epoch": 8.819488817891374, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 11042 + }, + { + "epoch": 8.820287539936102, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11043 + }, + { + "epoch": 8.821086261980831, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 11044 + }, + { + "epoch": 8.821884984025559, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 11045 + }, + { + "epoch": 8.822683706070288, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 11046 + }, + { + "epoch": 8.823482428115016, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 11047 + }, + { + "epoch": 8.824281150159745, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 11048 + }, + { + "epoch": 8.825079872204473, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 11049 + }, + { + "epoch": 8.8258785942492, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 11050 + }, + { + "epoch": 8.82667731629393, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11051 + }, + { + "epoch": 8.827476038338657, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 11052 + }, + { + "epoch": 8.828274760383387, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 11053 + }, + { + "epoch": 8.829073482428115, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 11054 + }, + { + "epoch": 8.829872204472844, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 11055 + }, + { + "epoch": 8.830670926517572, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 11056 + }, + { + "epoch": 8.831469648562301, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 11057 + }, + { + "epoch": 8.832268370607029, + "grad_norm": 0.091796875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 11058 + }, + { + "epoch": 8.833067092651756, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 11059 + }, + { + "epoch": 8.833865814696486, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 11060 + }, + { + "epoch": 8.834664536741213, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 11061 + }, + { + "epoch": 8.835463258785943, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 11062 + }, + { + "epoch": 8.83626198083067, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 11063 + }, + { + "epoch": 8.8370607028754, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 11064 + }, + { + "epoch": 8.837859424920127, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 11065 + }, + { + "epoch": 8.838658146964857, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 11066 + }, + { + "epoch": 8.839456869009584, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 11067 + }, + { + "epoch": 8.840255591054314, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 11068 + }, + { + "epoch": 8.841054313099042, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 11069 + }, + { + "epoch": 8.84185303514377, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 11070 + }, + { + "epoch": 8.842651757188499, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 11071 + }, + { + "epoch": 8.843450479233226, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 11072 + }, + { + "epoch": 8.844249201277956, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 11073 + }, + { + "epoch": 8.845047923322683, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 11074 + }, + { + "epoch": 8.845846645367413, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 11075 + }, + { + "epoch": 8.84664536741214, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11076 + }, + { + "epoch": 8.84744408945687, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 11077 + }, + { + "epoch": 8.848242811501597, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 11078 + }, + { + "epoch": 8.849041533546325, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11079 + }, + { + "epoch": 8.849840255591054, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 11080 + }, + { + "epoch": 8.850638977635782, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 11081 + }, + { + "epoch": 8.851437699680512, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11082 + }, + { + "epoch": 8.85223642172524, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 11083 + }, + { + "epoch": 8.853035143769969, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 11084 + }, + { + "epoch": 8.853833865814696, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 11085 + }, + { + "epoch": 8.854632587859426, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 11086 + }, + { + "epoch": 8.855431309904153, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 11087 + }, + { + "epoch": 8.856230031948883, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 11088 + }, + { + "epoch": 8.85702875399361, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 11089 + }, + { + "epoch": 8.857827476038338, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 11090 + }, + { + "epoch": 8.858626198083067, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 11091 + }, + { + "epoch": 8.859424920127795, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 11092 + }, + { + "epoch": 8.860223642172524, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 11093 + }, + { + "epoch": 8.861022364217252, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 11094 + }, + { + "epoch": 8.861821086261982, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 11095 + }, + { + "epoch": 8.86261980830671, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 11096 + }, + { + "epoch": 8.863418530351439, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 11097 + }, + { + "epoch": 8.864217252396166, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 11098 + }, + { + "epoch": 8.865015974440894, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 11099 + }, + { + "epoch": 8.865814696485623, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 11100 + }, + { + "epoch": 8.866613418530351, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 11101 + }, + { + "epoch": 8.86741214057508, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 11102 + }, + { + "epoch": 8.868210862619808, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 11103 + }, + { + "epoch": 8.869009584664537, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 11104 + }, + { + "epoch": 8.869808306709265, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 11105 + }, + { + "epoch": 8.870607028753994, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11106 + }, + { + "epoch": 8.871405750798722, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 11107 + }, + { + "epoch": 8.87220447284345, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11108 + }, + { + "epoch": 8.87300319488818, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 11109 + }, + { + "epoch": 8.873801916932907, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 11110 + }, + { + "epoch": 8.874600638977636, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 11111 + }, + { + "epoch": 8.875399361022364, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 11112 + }, + { + "epoch": 8.876198083067093, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 11113 + }, + { + "epoch": 8.87699680511182, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 11114 + }, + { + "epoch": 8.87779552715655, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 11115 + }, + { + "epoch": 8.878594249201278, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 11116 + }, + { + "epoch": 8.879392971246006, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 11117 + }, + { + "epoch": 8.880191693290735, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 11118 + }, + { + "epoch": 8.880990415335463, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 11119 + }, + { + "epoch": 8.881789137380192, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 11120 + }, + { + "epoch": 8.88258785942492, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 11121 + }, + { + "epoch": 8.883386581469649, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 11122 + }, + { + "epoch": 8.884185303514377, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 11123 + }, + { + "epoch": 8.884984025559106, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11124 + }, + { + "epoch": 8.885782747603834, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 11125 + }, + { + "epoch": 8.886581469648561, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 11126 + }, + { + "epoch": 8.88738019169329, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 11127 + }, + { + "epoch": 8.888178913738018, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 11128 + }, + { + "epoch": 8.888977635782748, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 11129 + }, + { + "epoch": 8.889776357827476, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 11130 + }, + { + "epoch": 8.890575079872205, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 11131 + }, + { + "epoch": 8.891373801916933, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 11132 + }, + { + "epoch": 8.892172523961662, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 11133 + }, + { + "epoch": 8.89297124600639, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 11134 + }, + { + "epoch": 8.893769968051117, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 11135 + }, + { + "epoch": 8.894568690095847, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 11136 + }, + { + "epoch": 8.895367412140574, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 11137 + }, + { + "epoch": 8.896166134185304, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 11138 + }, + { + "epoch": 8.896964856230031, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 11139 + }, + { + "epoch": 8.89776357827476, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 11140 + }, + { + "epoch": 8.898562300319488, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 11141 + }, + { + "epoch": 8.899361022364218, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 11142 + }, + { + "epoch": 8.900159744408946, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11143 + }, + { + "epoch": 8.900958466453675, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 11144 + }, + { + "epoch": 8.901757188498403, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 11145 + }, + { + "epoch": 8.90255591054313, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 11146 + }, + { + "epoch": 8.90335463258786, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 11147 + }, + { + "epoch": 8.904153354632587, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 11148 + }, + { + "epoch": 8.904952076677317, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 11149 + }, + { + "epoch": 8.905750798722044, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 11150 + }, + { + "epoch": 8.906549520766774, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 11151 + }, + { + "epoch": 8.907348242811501, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 11152 + }, + { + "epoch": 8.90814696485623, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 11153 + }, + { + "epoch": 8.908945686900958, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1152, + "step": 11154 + }, + { + "epoch": 8.909744408945686, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 11155 + }, + { + "epoch": 8.910543130990416, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11156 + }, + { + "epoch": 8.911341853035143, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 11157 + }, + { + "epoch": 8.912140575079873, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 11158 + }, + { + "epoch": 8.9129392971246, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 11159 + }, + { + "epoch": 8.91373801916933, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 11160 + }, + { + "epoch": 8.914536741214057, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 11161 + }, + { + "epoch": 8.915335463258787, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 11162 + }, + { + "epoch": 8.916134185303514, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 11163 + }, + { + "epoch": 8.916932907348244, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 11164 + }, + { + "epoch": 8.917731629392971, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 11165 + }, + { + "epoch": 8.918530351437699, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 11166 + }, + { + "epoch": 8.919329073482428, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 11167 + }, + { + "epoch": 8.920127795527156, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 11168 + }, + { + "epoch": 8.920926517571885, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 11169 + }, + { + "epoch": 8.921725239616613, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 11170 + }, + { + "epoch": 8.922523961661343, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 11171 + }, + { + "epoch": 8.92332268370607, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11172 + }, + { + "epoch": 8.9241214057508, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 11173 + }, + { + "epoch": 8.924920127795527, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 11174 + }, + { + "epoch": 8.925718849840255, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 11175 + }, + { + "epoch": 8.926517571884984, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11176 + }, + { + "epoch": 8.927316293929712, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 11177 + }, + { + "epoch": 8.928115015974441, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 11178 + }, + { + "epoch": 8.928913738019169, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 11179 + }, + { + "epoch": 8.929712460063898, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11180 + }, + { + "epoch": 8.930511182108626, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 11181 + }, + { + "epoch": 8.931309904153355, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 11182 + }, + { + "epoch": 8.932108626198083, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 11183 + }, + { + "epoch": 8.93290734824281, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 11184 + }, + { + "epoch": 8.93370607028754, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 11185 + }, + { + "epoch": 8.934504792332268, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 11186 + }, + { + "epoch": 8.935303514376997, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11187 + }, + { + "epoch": 8.936102236421725, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 11188 + }, + { + "epoch": 8.936900958466454, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 11189 + }, + { + "epoch": 8.937699680511182, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 11190 + }, + { + "epoch": 8.938498402555911, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 11191 + }, + { + "epoch": 8.939297124600639, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 11192 + }, + { + "epoch": 8.940095846645367, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 11193 + }, + { + "epoch": 8.940894568690096, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 11194 + }, + { + "epoch": 8.941693290734824, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 11195 + }, + { + "epoch": 8.942492012779553, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 11196 + }, + { + "epoch": 8.94329073482428, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 11197 + }, + { + "epoch": 8.94408945686901, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 11198 + }, + { + "epoch": 8.944888178913738, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 11199 + }, + { + "epoch": 8.945686900958467, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11200 + }, + { + "epoch": 8.946485623003195, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11201 + }, + { + "epoch": 8.947284345047922, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 11202 + }, + { + "epoch": 8.948083067092652, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 11203 + }, + { + "epoch": 8.94888178913738, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 11204 + }, + { + "epoch": 8.949680511182109, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 11205 + }, + { + "epoch": 8.950479233226837, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 11206 + }, + { + "epoch": 8.951277955271566, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 11207 + }, + { + "epoch": 8.952076677316294, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 11208 + }, + { + "epoch": 8.952875399361023, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 11209 + }, + { + "epoch": 8.95367412140575, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 11210 + }, + { + "epoch": 8.954472843450478, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 11211 + }, + { + "epoch": 8.955271565495208, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 11212 + }, + { + "epoch": 8.956070287539935, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1174, + "step": 11213 + }, + { + "epoch": 8.956869009584665, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 11214 + }, + { + "epoch": 8.957667731629392, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 11215 + }, + { + "epoch": 8.958466453674122, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 11216 + }, + { + "epoch": 8.95926517571885, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 11217 + }, + { + "epoch": 8.960063897763579, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 11218 + }, + { + "epoch": 8.960862619808307, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 11219 + }, + { + "epoch": 8.961661341853034, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 11220 + }, + { + "epoch": 8.962460063897764, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 11221 + }, + { + "epoch": 8.963258785942491, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 11222 + }, + { + "epoch": 8.96405750798722, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 11223 + }, + { + "epoch": 8.964856230031948, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 11224 + }, + { + "epoch": 8.965654952076678, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 11225 + }, + { + "epoch": 8.966453674121405, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 11226 + }, + { + "epoch": 8.967252396166135, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 11227 + }, + { + "epoch": 8.968051118210862, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 11228 + }, + { + "epoch": 8.968849840255592, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 11229 + }, + { + "epoch": 8.96964856230032, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 11230 + }, + { + "epoch": 8.970447284345047, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 11231 + }, + { + "epoch": 8.971246006389777, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 11232 + }, + { + "epoch": 8.972044728434504, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 11233 + }, + { + "epoch": 8.972843450479234, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 11234 + }, + { + "epoch": 8.973642172523961, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 11235 + }, + { + "epoch": 8.97444089456869, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 11236 + }, + { + "epoch": 8.975239616613418, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11237 + }, + { + "epoch": 8.976038338658148, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 11238 + }, + { + "epoch": 8.976837060702875, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 11239 + }, + { + "epoch": 8.977635782747605, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 11240 + }, + { + "epoch": 8.978434504792332, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 11241 + }, + { + "epoch": 8.97923322683706, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 11242 + }, + { + "epoch": 8.98003194888179, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 11243 + }, + { + "epoch": 8.980830670926517, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 11244 + }, + { + "epoch": 8.981629392971247, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11245 + }, + { + "epoch": 8.982428115015974, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 11246 + }, + { + "epoch": 8.983226837060704, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 11247 + }, + { + "epoch": 8.984025559105431, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 11248 + }, + { + "epoch": 8.98482428115016, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 11249 + }, + { + "epoch": 8.985623003194888, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 11250 + }, + { + "epoch": 8.986421725239616, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 11251 + }, + { + "epoch": 8.987220447284345, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 11252 + }, + { + "epoch": 8.988019169329073, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 11253 + }, + { + "epoch": 8.988817891373802, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 11254 + }, + { + "epoch": 8.98961661341853, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 11255 + }, + { + "epoch": 8.99041533546326, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 11256 + }, + { + "epoch": 8.991214057507987, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11257 + }, + { + "epoch": 8.992012779552716, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 11258 + }, + { + "epoch": 8.992811501597444, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11259 + }, + { + "epoch": 8.993610223642172, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 11260 + }, + { + "epoch": 8.994408945686901, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 11261 + }, + { + "epoch": 8.995207667731629, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 11262 + }, + { + "epoch": 8.996006389776358, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 11263 + }, + { + "epoch": 8.996805111821086, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 11264 + }, + { + "epoch": 8.997603833865815, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 11265 + }, + { + "epoch": 8.998402555910543, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 11266 + }, + { + "epoch": 8.999201277955272, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 11267 + }, + { + "epoch": 9.0, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 11268 + }, + { + "epoch": 9.000798722044728, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 11269 + }, + { + "epoch": 9.001597444089457, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 11270 + }, + { + "epoch": 9.002396166134185, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 11271 + }, + { + "epoch": 9.003194888178914, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 11272 + }, + { + "epoch": 9.003993610223642, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 11273 + }, + { + "epoch": 9.004792332268371, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 11274 + }, + { + "epoch": 9.005591054313099, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 11275 + }, + { + "epoch": 9.006389776357828, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11276 + }, + { + "epoch": 9.007188498402556, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 11277 + }, + { + "epoch": 9.007987220447284, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 11278 + }, + { + "epoch": 9.008785942492013, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 11279 + }, + { + "epoch": 9.00958466453674, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 11280 + }, + { + "epoch": 9.01038338658147, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 11281 + }, + { + "epoch": 9.011182108626198, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 11282 + }, + { + "epoch": 9.011980830670927, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 11283 + }, + { + "epoch": 9.012779552715655, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 11284 + }, + { + "epoch": 9.013578274760384, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11285 + }, + { + "epoch": 9.014376996805112, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 11286 + }, + { + "epoch": 9.01517571884984, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 11287 + }, + { + "epoch": 9.015974440894569, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11288 + }, + { + "epoch": 9.016773162939296, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 11289 + }, + { + "epoch": 9.017571884984026, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 11290 + }, + { + "epoch": 9.018370607028753, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 11291 + }, + { + "epoch": 9.019169329073483, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 11292 + }, + { + "epoch": 9.01996805111821, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 11293 + }, + { + "epoch": 9.02076677316294, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 11294 + }, + { + "epoch": 9.021565495207668, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 11295 + }, + { + "epoch": 9.022364217252397, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 11296 + }, + { + "epoch": 9.023162939297125, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 11297 + }, + { + "epoch": 9.023961661341852, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 11298 + }, + { + "epoch": 9.024760383386582, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11299 + }, + { + "epoch": 9.02555910543131, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 11300 + }, + { + "epoch": 9.026357827476039, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11301 + }, + { + "epoch": 9.027156549520766, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 11302 + }, + { + "epoch": 9.027955271565496, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 11303 + }, + { + "epoch": 9.028753993610223, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 11304 + }, + { + "epoch": 9.029552715654953, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 11305 + }, + { + "epoch": 9.03035143769968, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11306 + }, + { + "epoch": 9.031150159744408, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 11307 + }, + { + "epoch": 9.031948881789138, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 11308 + }, + { + "epoch": 9.032747603833865, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 11309 + }, + { + "epoch": 9.033546325878595, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 11310 + }, + { + "epoch": 9.034345047923322, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 11311 + }, + { + "epoch": 9.035143769968052, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 11312 + }, + { + "epoch": 9.03594249201278, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 11313 + }, + { + "epoch": 9.036741214057509, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 11314 + }, + { + "epoch": 9.037539936102236, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 11315 + }, + { + "epoch": 9.038338658146964, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 11316 + }, + { + "epoch": 9.039137380191693, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 11317 + }, + { + "epoch": 9.039936102236421, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 11318 + }, + { + "epoch": 9.04073482428115, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 11319 + }, + { + "epoch": 9.041533546325878, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11320 + }, + { + "epoch": 9.042332268370608, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 11321 + }, + { + "epoch": 9.043130990415335, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 11322 + }, + { + "epoch": 9.043929712460065, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 11323 + }, + { + "epoch": 9.044728434504792, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 11324 + }, + { + "epoch": 9.04552715654952, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11325 + }, + { + "epoch": 9.04632587859425, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 11326 + }, + { + "epoch": 9.047124600638977, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 11327 + }, + { + "epoch": 9.047923322683706, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 11328 + }, + { + "epoch": 9.048722044728434, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 11329 + }, + { + "epoch": 9.049520766773163, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 11330 + }, + { + "epoch": 9.050319488817891, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 11331 + }, + { + "epoch": 9.05111821086262, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 11332 + }, + { + "epoch": 9.051916932907348, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 11333 + }, + { + "epoch": 9.052715654952078, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 11334 + }, + { + "epoch": 9.053514376996805, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 11335 + }, + { + "epoch": 9.054313099041533, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 11336 + }, + { + "epoch": 9.055111821086262, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 11337 + }, + { + "epoch": 9.05591054313099, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 11338 + }, + { + "epoch": 9.05670926517572, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 11339 + }, + { + "epoch": 9.057507987220447, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 11340 + }, + { + "epoch": 9.058306709265176, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11341 + }, + { + "epoch": 9.059105431309904, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 11342 + }, + { + "epoch": 9.059904153354633, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 11343 + }, + { + "epoch": 9.060702875399361, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 11344 + }, + { + "epoch": 9.061501597444089, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 11345 + }, + { + "epoch": 9.062300319488818, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 11346 + }, + { + "epoch": 9.063099041533546, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 11347 + }, + { + "epoch": 9.063897763578275, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 11348 + }, + { + "epoch": 9.064696485623003, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 11349 + }, + { + "epoch": 9.065495207667732, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 11350 + }, + { + "epoch": 9.06629392971246, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 11351 + }, + { + "epoch": 9.06709265175719, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 11352 + }, + { + "epoch": 9.067891373801917, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 11353 + }, + { + "epoch": 9.068690095846645, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11354 + }, + { + "epoch": 9.069488817891374, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 11355 + }, + { + "epoch": 9.070287539936102, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 11356 + }, + { + "epoch": 9.071086261980831, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 11357 + }, + { + "epoch": 9.071884984025559, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 11358 + }, + { + "epoch": 9.072683706070288, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 11359 + }, + { + "epoch": 9.073482428115016, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 11360 + }, + { + "epoch": 9.074281150159745, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 11361 + }, + { + "epoch": 9.075079872204473, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 11362 + }, + { + "epoch": 9.0758785942492, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 11363 + }, + { + "epoch": 9.07667731629393, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 11364 + }, + { + "epoch": 9.077476038338657, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 11365 + }, + { + "epoch": 9.078274760383387, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 11366 + }, + { + "epoch": 9.079073482428115, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11367 + }, + { + "epoch": 9.079872204472844, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 11368 + }, + { + "epoch": 9.080670926517572, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 11369 + }, + { + "epoch": 9.081469648562301, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 11370 + }, + { + "epoch": 9.082268370607029, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 11371 + }, + { + "epoch": 9.083067092651758, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 11372 + }, + { + "epoch": 9.083865814696486, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 11373 + }, + { + "epoch": 9.084664536741213, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 11374 + }, + { + "epoch": 9.085463258785943, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 11375 + }, + { + "epoch": 9.08626198083067, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 11376 + }, + { + "epoch": 9.0870607028754, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 11377 + }, + { + "epoch": 9.087859424920127, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 11378 + }, + { + "epoch": 9.088658146964857, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 11379 + }, + { + "epoch": 9.089456869009584, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 11380 + }, + { + "epoch": 9.090255591054314, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 11381 + }, + { + "epoch": 9.091054313099042, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 11382 + }, + { + "epoch": 9.09185303514377, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 11383 + }, + { + "epoch": 9.092651757188499, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 11384 + }, + { + "epoch": 9.093450479233226, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 11385 + }, + { + "epoch": 9.094249201277956, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 11386 + }, + { + "epoch": 9.095047923322683, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 11387 + }, + { + "epoch": 9.095846645367413, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 11388 + }, + { + "epoch": 9.09664536741214, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 11389 + }, + { + "epoch": 9.09744408945687, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 11390 + }, + { + "epoch": 9.098242811501597, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 11391 + }, + { + "epoch": 9.099041533546325, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 11392 + }, + { + "epoch": 9.099840255591054, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 11393 + }, + { + "epoch": 9.100638977635782, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 11394 + }, + { + "epoch": 9.101437699680512, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 11395 + }, + { + "epoch": 9.10223642172524, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 11396 + }, + { + "epoch": 9.103035143769969, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 11397 + }, + { + "epoch": 9.103833865814696, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 11398 + }, + { + "epoch": 9.104632587859426, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11399 + }, + { + "epoch": 9.105431309904153, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 11400 + }, + { + "epoch": 9.106230031948881, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 11401 + }, + { + "epoch": 9.10702875399361, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 11402 + }, + { + "epoch": 9.107827476038338, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 11403 + }, + { + "epoch": 9.108626198083067, + "grad_norm": 0.09765625, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 11404 + }, + { + "epoch": 9.109424920127795, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 11405 + }, + { + "epoch": 9.110223642172524, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 11406 + }, + { + "epoch": 9.111022364217252, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 11407 + }, + { + "epoch": 9.111821086261982, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 11408 + }, + { + "epoch": 9.11261980830671, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11409 + }, + { + "epoch": 9.113418530351439, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 11410 + }, + { + "epoch": 9.114217252396166, + "grad_norm": 0.09375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 11411 + }, + { + "epoch": 9.115015974440894, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 11412 + }, + { + "epoch": 9.115814696485623, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 11413 + }, + { + "epoch": 9.116613418530351, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 11414 + }, + { + "epoch": 9.11741214057508, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 11415 + }, + { + "epoch": 9.118210862619808, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 11416 + }, + { + "epoch": 9.119009584664537, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 11417 + }, + { + "epoch": 9.119808306709265, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 11418 + }, + { + "epoch": 9.120607028753994, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 11419 + }, + { + "epoch": 9.121405750798722, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 11420 + }, + { + "epoch": 9.12220447284345, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 11421 + }, + { + "epoch": 9.12300319488818, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 11422 + }, + { + "epoch": 9.123801916932907, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 11423 + }, + { + "epoch": 9.124600638977636, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 11424 + }, + { + "epoch": 9.125399361022364, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 11425 + }, + { + "epoch": 9.126198083067093, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 11426 + }, + { + "epoch": 9.12699680511182, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11427 + }, + { + "epoch": 9.12779552715655, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 11428 + }, + { + "epoch": 9.128594249201278, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 11429 + }, + { + "epoch": 9.129392971246006, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 11430 + }, + { + "epoch": 9.130191693290735, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 11431 + }, + { + "epoch": 9.130990415335463, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 11432 + }, + { + "epoch": 9.131789137380192, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 11433 + }, + { + "epoch": 9.13258785942492, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 11434 + }, + { + "epoch": 9.133386581469649, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 11435 + }, + { + "epoch": 9.134185303514377, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 11436 + }, + { + "epoch": 9.134984025559106, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 11437 + }, + { + "epoch": 9.135782747603834, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 11438 + }, + { + "epoch": 9.136581469648561, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 11439 + }, + { + "epoch": 9.13738019169329, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 11440 + }, + { + "epoch": 9.138178913738018, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 11441 + }, + { + "epoch": 9.138977635782748, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 11442 + }, + { + "epoch": 9.139776357827476, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 11443 + }, + { + "epoch": 9.140575079872205, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 11444 + }, + { + "epoch": 9.141373801916933, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 11445 + }, + { + "epoch": 9.142172523961662, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 11446 + }, + { + "epoch": 9.14297124600639, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 11447 + }, + { + "epoch": 9.143769968051119, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11448 + }, + { + "epoch": 9.144568690095847, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 11449 + }, + { + "epoch": 9.145367412140574, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 11450 + }, + { + "epoch": 9.146166134185304, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 11451 + }, + { + "epoch": 9.146964856230031, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 11452 + }, + { + "epoch": 9.14776357827476, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 11453 + }, + { + "epoch": 9.148562300319488, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 11454 + }, + { + "epoch": 9.149361022364218, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11455 + }, + { + "epoch": 9.150159744408946, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 11456 + }, + { + "epoch": 9.150958466453675, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 11457 + }, + { + "epoch": 9.151757188498403, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11458 + }, + { + "epoch": 9.15255591054313, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 11459 + }, + { + "epoch": 9.15335463258786, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 11460 + }, + { + "epoch": 9.154153354632587, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 11461 + }, + { + "epoch": 9.154952076677317, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 11462 + }, + { + "epoch": 9.155750798722044, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 11463 + }, + { + "epoch": 9.156549520766774, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 11464 + }, + { + "epoch": 9.157348242811501, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 11465 + }, + { + "epoch": 9.15814696485623, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 11466 + }, + { + "epoch": 9.158945686900958, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 11467 + }, + { + "epoch": 9.159744408945686, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 11468 + }, + { + "epoch": 9.160543130990416, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 11469 + }, + { + "epoch": 9.161341853035143, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11470 + }, + { + "epoch": 9.162140575079873, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 11471 + }, + { + "epoch": 9.1629392971246, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 11472 + }, + { + "epoch": 9.16373801916933, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11473 + }, + { + "epoch": 9.164536741214057, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 11474 + }, + { + "epoch": 9.165335463258787, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 11475 + }, + { + "epoch": 9.166134185303514, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 11476 + }, + { + "epoch": 9.166932907348242, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 11477 + }, + { + "epoch": 9.167731629392971, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11478 + }, + { + "epoch": 9.168530351437699, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 11479 + }, + { + "epoch": 9.169329073482428, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 11480 + }, + { + "epoch": 9.170127795527156, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 11481 + }, + { + "epoch": 9.170926517571885, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 11482 + }, + { + "epoch": 9.171725239616613, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 11483 + }, + { + "epoch": 9.172523961661343, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 11484 + }, + { + "epoch": 9.17332268370607, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 11485 + }, + { + "epoch": 9.1741214057508, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 11486 + }, + { + "epoch": 9.174920127795527, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 11487 + }, + { + "epoch": 9.175718849840255, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 11488 + }, + { + "epoch": 9.176517571884984, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11489 + }, + { + "epoch": 9.177316293929712, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11490 + }, + { + "epoch": 9.178115015974441, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 11491 + }, + { + "epoch": 9.178913738019169, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 11492 + }, + { + "epoch": 9.179712460063898, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 11493 + }, + { + "epoch": 9.180511182108626, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11494 + }, + { + "epoch": 9.181309904153355, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 11495 + }, + { + "epoch": 9.182108626198083, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 11496 + }, + { + "epoch": 9.18290734824281, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 11497 + }, + { + "epoch": 9.18370607028754, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 11498 + }, + { + "epoch": 9.184504792332268, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 11499 + }, + { + "epoch": 9.185303514376997, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 11500 + }, + { + "epoch": 9.186102236421725, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 11501 + }, + { + "epoch": 9.186900958466454, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 11502 + }, + { + "epoch": 9.187699680511182, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 11503 + }, + { + "epoch": 9.188498402555911, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 11504 + }, + { + "epoch": 9.189297124600639, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 11505 + }, + { + "epoch": 9.190095846645367, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 11506 + }, + { + "epoch": 9.190894568690096, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 11507 + }, + { + "epoch": 9.191693290734824, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 11508 + }, + { + "epoch": 9.192492012779553, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 11509 + }, + { + "epoch": 9.19329073482428, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 11510 + }, + { + "epoch": 9.19408945686901, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 11511 + }, + { + "epoch": 9.194888178913738, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 11512 + }, + { + "epoch": 9.195686900958467, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 11513 + }, + { + "epoch": 9.196485623003195, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 11514 + }, + { + "epoch": 9.197284345047922, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 11515 + }, + { + "epoch": 9.198083067092652, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 11516 + }, + { + "epoch": 9.19888178913738, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 11517 + }, + { + "epoch": 9.199680511182109, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 11518 + }, + { + "epoch": 9.200479233226837, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 11519 + }, + { + "epoch": 9.201277955271566, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 11520 + }, + { + "epoch": 9.202076677316294, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 11521 + }, + { + "epoch": 9.202875399361023, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 11522 + }, + { + "epoch": 9.20367412140575, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 11523 + }, + { + "epoch": 9.204472843450478, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 11524 + }, + { + "epoch": 9.205271565495208, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 11525 + }, + { + "epoch": 9.206070287539935, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 11526 + }, + { + "epoch": 9.206869009584665, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 11527 + }, + { + "epoch": 9.207667731629392, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 11528 + }, + { + "epoch": 9.208466453674122, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 11529 + }, + { + "epoch": 9.20926517571885, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 11530 + }, + { + "epoch": 9.210063897763579, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 11531 + }, + { + "epoch": 9.210862619808307, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 11532 + }, + { + "epoch": 9.211661341853036, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 11533 + }, + { + "epoch": 9.212460063897764, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1136, + "step": 11534 + }, + { + "epoch": 9.213258785942491, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 11535 + }, + { + "epoch": 9.21405750798722, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 11536 + }, + { + "epoch": 9.214856230031948, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 11537 + }, + { + "epoch": 9.215654952076678, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11538 + }, + { + "epoch": 9.216453674121405, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 11539 + }, + { + "epoch": 9.217252396166135, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 11540 + }, + { + "epoch": 9.218051118210862, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 11541 + }, + { + "epoch": 9.218849840255592, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 11542 + }, + { + "epoch": 9.21964856230032, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 11543 + }, + { + "epoch": 9.220447284345047, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11544 + }, + { + "epoch": 9.221246006389777, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 11545 + }, + { + "epoch": 9.222044728434504, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 11546 + }, + { + "epoch": 9.222843450479234, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 11547 + }, + { + "epoch": 9.223642172523961, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 11548 + }, + { + "epoch": 9.22444089456869, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 11549 + }, + { + "epoch": 9.225239616613418, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 11550 + }, + { + "epoch": 9.226038338658148, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 11551 + }, + { + "epoch": 9.226837060702875, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 11552 + }, + { + "epoch": 9.227635782747603, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 11553 + }, + { + "epoch": 9.228434504792332, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 11554 + }, + { + "epoch": 9.22923322683706, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 11555 + }, + { + "epoch": 9.23003194888179, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11556 + }, + { + "epoch": 9.230830670926517, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 11557 + }, + { + "epoch": 9.231629392971247, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 11558 + }, + { + "epoch": 9.232428115015974, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 11559 + }, + { + "epoch": 9.233226837060704, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 11560 + }, + { + "epoch": 9.234025559105431, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 11561 + }, + { + "epoch": 9.23482428115016, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 11562 + }, + { + "epoch": 9.235623003194888, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 11563 + }, + { + "epoch": 9.236421725239616, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 11564 + }, + { + "epoch": 9.237220447284345, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 11565 + }, + { + "epoch": 9.238019169329073, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 11566 + }, + { + "epoch": 9.238817891373802, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 11567 + }, + { + "epoch": 9.23961661341853, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 11568 + }, + { + "epoch": 9.24041533546326, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 11569 + }, + { + "epoch": 9.241214057507987, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 11570 + }, + { + "epoch": 9.242012779552716, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 11571 + }, + { + "epoch": 9.242811501597444, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 11572 + }, + { + "epoch": 9.243610223642172, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 11573 + }, + { + "epoch": 9.244408945686901, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 11574 + }, + { + "epoch": 9.245207667731629, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 11575 + }, + { + "epoch": 9.246006389776358, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11576 + }, + { + "epoch": 9.246805111821086, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 11577 + }, + { + "epoch": 9.247603833865815, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 11578 + }, + { + "epoch": 9.248402555910543, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 11579 + }, + { + "epoch": 9.249201277955272, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 11580 + }, + { + "epoch": 9.25, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 11581 + }, + { + "epoch": 9.250798722044728, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 11582 + }, + { + "epoch": 9.251597444089457, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11583 + }, + { + "epoch": 9.252396166134185, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 11584 + }, + { + "epoch": 9.253194888178914, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 11585 + }, + { + "epoch": 9.253993610223642, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 11586 + }, + { + "epoch": 9.254792332268371, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 11587 + }, + { + "epoch": 9.255591054313099, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 11588 + }, + { + "epoch": 9.256389776357828, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 11589 + }, + { + "epoch": 9.257188498402556, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 11590 + }, + { + "epoch": 9.257987220447284, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 11591 + }, + { + "epoch": 9.258785942492013, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 11592 + }, + { + "epoch": 9.25958466453674, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 11593 + }, + { + "epoch": 9.26038338658147, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 11594 + }, + { + "epoch": 9.261182108626198, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 11595 + }, + { + "epoch": 9.261980830670927, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 11596 + }, + { + "epoch": 9.262779552715655, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 11597 + }, + { + "epoch": 9.263578274760384, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 11598 + }, + { + "epoch": 9.264376996805112, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 11599 + }, + { + "epoch": 9.26517571884984, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 11600 + }, + { + "epoch": 9.265974440894569, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 11601 + }, + { + "epoch": 9.266773162939296, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 11602 + }, + { + "epoch": 9.267571884984026, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 11603 + }, + { + "epoch": 9.268370607028753, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 11604 + }, + { + "epoch": 9.269169329073483, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11605 + }, + { + "epoch": 9.26996805111821, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 11606 + }, + { + "epoch": 9.27076677316294, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 11607 + }, + { + "epoch": 9.271565495207668, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 11608 + }, + { + "epoch": 9.272364217252397, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 11609 + }, + { + "epoch": 9.273162939297125, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 11610 + }, + { + "epoch": 9.273961661341852, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 11611 + }, + { + "epoch": 9.274760383386582, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 11612 + }, + { + "epoch": 9.27555910543131, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 11613 + }, + { + "epoch": 9.276357827476039, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11614 + }, + { + "epoch": 9.277156549520766, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 11615 + }, + { + "epoch": 9.277955271565496, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 11616 + }, + { + "epoch": 9.278753993610223, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11617 + }, + { + "epoch": 9.279552715654953, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 11618 + }, + { + "epoch": 9.28035143769968, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 11619 + }, + { + "epoch": 9.281150159744408, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 11620 + }, + { + "epoch": 9.281948881789138, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 11621 + }, + { + "epoch": 9.282747603833865, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 11622 + }, + { + "epoch": 9.283546325878595, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 11623 + }, + { + "epoch": 9.284345047923322, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 11624 + }, + { + "epoch": 9.285143769968052, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11625 + }, + { + "epoch": 9.28594249201278, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 11626 + }, + { + "epoch": 9.286741214057509, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 11627 + }, + { + "epoch": 9.287539936102236, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 11628 + }, + { + "epoch": 9.288338658146964, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 11629 + }, + { + "epoch": 9.289137380191693, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 11630 + }, + { + "epoch": 9.289936102236421, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 11631 + }, + { + "epoch": 9.29073482428115, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 11632 + }, + { + "epoch": 9.291533546325878, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 11633 + }, + { + "epoch": 9.292332268370608, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 11634 + }, + { + "epoch": 9.293130990415335, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 11635 + }, + { + "epoch": 9.293929712460065, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 11636 + }, + { + "epoch": 9.294728434504792, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 11637 + }, + { + "epoch": 9.295527156549522, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 11638 + }, + { + "epoch": 9.29632587859425, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 11639 + }, + { + "epoch": 9.297124600638977, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 11640 + }, + { + "epoch": 9.297923322683706, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 11641 + }, + { + "epoch": 9.298722044728434, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 11642 + }, + { + "epoch": 9.299520766773163, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 11643 + }, + { + "epoch": 9.300319488817891, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 11644 + }, + { + "epoch": 9.30111821086262, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 11645 + }, + { + "epoch": 9.301916932907348, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 11646 + }, + { + "epoch": 9.302715654952078, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 11647 + }, + { + "epoch": 9.303514376996805, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 11648 + }, + { + "epoch": 9.304313099041533, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 11649 + }, + { + "epoch": 9.305111821086262, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 11650 + }, + { + "epoch": 9.30591054313099, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 11651 + }, + { + "epoch": 9.30670926517572, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 11652 + }, + { + "epoch": 9.307507987220447, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 11653 + }, + { + "epoch": 9.308306709265176, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 11654 + }, + { + "epoch": 9.309105431309904, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 11655 + }, + { + "epoch": 9.309904153354633, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11656 + }, + { + "epoch": 9.310702875399361, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 11657 + }, + { + "epoch": 9.311501597444089, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 11658 + }, + { + "epoch": 9.312300319488818, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 11659 + }, + { + "epoch": 9.313099041533546, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 11660 + }, + { + "epoch": 9.313897763578275, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 11661 + }, + { + "epoch": 9.314696485623003, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 11662 + }, + { + "epoch": 9.315495207667732, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 11663 + }, + { + "epoch": 9.31629392971246, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 11664 + }, + { + "epoch": 9.31709265175719, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 11665 + }, + { + "epoch": 9.317891373801917, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 11666 + }, + { + "epoch": 9.318690095846645, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 11667 + }, + { + "epoch": 9.319488817891374, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 11668 + }, + { + "epoch": 9.320287539936102, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 11669 + }, + { + "epoch": 9.321086261980831, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 11670 + }, + { + "epoch": 9.321884984025559, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 11671 + }, + { + "epoch": 9.322683706070288, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 11672 + }, + { + "epoch": 9.323482428115016, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 11673 + }, + { + "epoch": 9.324281150159745, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 11674 + }, + { + "epoch": 9.325079872204473, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 11675 + }, + { + "epoch": 9.3258785942492, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 11676 + }, + { + "epoch": 9.32667731629393, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 11677 + }, + { + "epoch": 9.327476038338657, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 11678 + }, + { + "epoch": 9.328274760383387, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 11679 + }, + { + "epoch": 9.329073482428115, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 11680 + }, + { + "epoch": 9.329872204472844, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 11681 + }, + { + "epoch": 9.330670926517572, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 11682 + }, + { + "epoch": 9.331469648562301, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 11683 + }, + { + "epoch": 9.332268370607029, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 11684 + }, + { + "epoch": 9.333067092651758, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 11685 + }, + { + "epoch": 9.333865814696486, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 11686 + }, + { + "epoch": 9.334664536741213, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 11687 + }, + { + "epoch": 9.335463258785943, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 11688 + }, + { + "epoch": 9.33626198083067, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 11689 + }, + { + "epoch": 9.3370607028754, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11690 + }, + { + "epoch": 9.337859424920127, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 11691 + }, + { + "epoch": 9.338658146964857, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 11692 + }, + { + "epoch": 9.339456869009584, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 11693 + }, + { + "epoch": 9.340255591054314, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 11694 + }, + { + "epoch": 9.341054313099042, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 11695 + }, + { + "epoch": 9.34185303514377, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 11696 + }, + { + "epoch": 9.342651757188499, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11697 + }, + { + "epoch": 9.343450479233226, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 11698 + }, + { + "epoch": 9.344249201277956, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 11699 + }, + { + "epoch": 9.345047923322683, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 11700 + }, + { + "epoch": 9.345846645367413, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 11701 + }, + { + "epoch": 9.34664536741214, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 11702 + }, + { + "epoch": 9.34744408945687, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 11703 + }, + { + "epoch": 9.348242811501597, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 11704 + }, + { + "epoch": 9.349041533546325, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 11705 + }, + { + "epoch": 9.349840255591054, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 11706 + }, + { + "epoch": 9.350638977635782, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 11707 + }, + { + "epoch": 9.351437699680512, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 11708 + }, + { + "epoch": 9.35223642172524, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 11709 + }, + { + "epoch": 9.353035143769969, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 11710 + }, + { + "epoch": 9.353833865814696, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 11711 + }, + { + "epoch": 9.354632587859426, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 11712 + }, + { + "epoch": 9.355431309904153, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 11713 + }, + { + "epoch": 9.356230031948883, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 11714 + }, + { + "epoch": 9.35702875399361, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 11715 + }, + { + "epoch": 9.357827476038338, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 11716 + }, + { + "epoch": 9.358626198083067, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 11717 + }, + { + "epoch": 9.359424920127795, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 11718 + }, + { + "epoch": 9.360223642172524, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 11719 + }, + { + "epoch": 9.361022364217252, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 11720 + }, + { + "epoch": 9.361821086261982, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 11721 + }, + { + "epoch": 9.36261980830671, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 11722 + }, + { + "epoch": 9.363418530351439, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 11723 + }, + { + "epoch": 9.364217252396166, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 11724 + }, + { + "epoch": 9.365015974440894, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 11725 + }, + { + "epoch": 9.365814696485623, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 11726 + }, + { + "epoch": 9.366613418530351, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 11727 + }, + { + "epoch": 9.36741214057508, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 11728 + }, + { + "epoch": 9.368210862619808, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 11729 + }, + { + "epoch": 9.369009584664537, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11730 + }, + { + "epoch": 9.369808306709265, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 11731 + }, + { + "epoch": 9.370607028753994, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 11732 + }, + { + "epoch": 9.371405750798722, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 11733 + }, + { + "epoch": 9.37220447284345, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 11734 + }, + { + "epoch": 9.37300319488818, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 11735 + }, + { + "epoch": 9.373801916932907, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 11736 + }, + { + "epoch": 9.374600638977636, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 11737 + }, + { + "epoch": 9.375399361022364, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 11738 + }, + { + "epoch": 9.376198083067093, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 11739 + }, + { + "epoch": 9.37699680511182, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 11740 + }, + { + "epoch": 9.37779552715655, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 11741 + }, + { + "epoch": 9.378594249201278, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 11742 + }, + { + "epoch": 9.379392971246006, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 11743 + }, + { + "epoch": 9.380191693290735, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11744 + }, + { + "epoch": 9.380990415335463, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 11745 + }, + { + "epoch": 9.381789137380192, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 11746 + }, + { + "epoch": 9.38258785942492, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1144, + "step": 11747 + }, + { + "epoch": 9.383386581469649, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 11748 + }, + { + "epoch": 9.384185303514377, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 11749 + }, + { + "epoch": 9.384984025559106, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 11750 + }, + { + "epoch": 9.385782747603834, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 11751 + }, + { + "epoch": 9.386581469648561, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 11752 + }, + { + "epoch": 9.38738019169329, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 11753 + }, + { + "epoch": 9.388178913738018, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 11754 + }, + { + "epoch": 9.388977635782748, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 11755 + }, + { + "epoch": 9.389776357827476, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 11756 + }, + { + "epoch": 9.390575079872205, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 11757 + }, + { + "epoch": 9.391373801916933, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 11758 + }, + { + "epoch": 9.392172523961662, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 11759 + }, + { + "epoch": 9.39297124600639, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 11760 + }, + { + "epoch": 9.393769968051117, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 11761 + }, + { + "epoch": 9.394568690095847, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 11762 + }, + { + "epoch": 9.395367412140574, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 11763 + }, + { + "epoch": 9.396166134185304, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 11764 + }, + { + "epoch": 9.396964856230031, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 11765 + }, + { + "epoch": 9.39776357827476, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 11766 + }, + { + "epoch": 9.398562300319488, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 11767 + }, + { + "epoch": 9.399361022364218, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 11768 + }, + { + "epoch": 9.400159744408946, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 11769 + }, + { + "epoch": 9.400958466453675, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 11770 + }, + { + "epoch": 9.401757188498403, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 11771 + }, + { + "epoch": 9.40255591054313, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 11772 + }, + { + "epoch": 9.40335463258786, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 11773 + }, + { + "epoch": 9.404153354632587, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 11774 + }, + { + "epoch": 9.404952076677317, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 11775 + }, + { + "epoch": 9.405750798722044, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 11776 + }, + { + "epoch": 9.406549520766774, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 11777 + }, + { + "epoch": 9.407348242811501, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 11778 + }, + { + "epoch": 9.40814696485623, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 11779 + }, + { + "epoch": 9.408945686900958, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 11780 + }, + { + "epoch": 9.409744408945686, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 11781 + }, + { + "epoch": 9.410543130990416, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 11782 + }, + { + "epoch": 9.411341853035143, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 11783 + }, + { + "epoch": 9.412140575079873, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 11784 + }, + { + "epoch": 9.4129392971246, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 11785 + }, + { + "epoch": 9.41373801916933, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 11786 + }, + { + "epoch": 9.414536741214057, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 11787 + }, + { + "epoch": 9.415335463258787, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 11788 + }, + { + "epoch": 9.416134185303514, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 11789 + }, + { + "epoch": 9.416932907348242, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 11790 + }, + { + "epoch": 9.417731629392971, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 11791 + }, + { + "epoch": 9.418530351437699, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 11792 + }, + { + "epoch": 9.419329073482428, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 11793 + }, + { + "epoch": 9.420127795527156, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 11794 + }, + { + "epoch": 9.420926517571885, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 11795 + }, + { + "epoch": 9.421725239616613, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 11796 + }, + { + "epoch": 9.422523961661343, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 11797 + }, + { + "epoch": 9.42332268370607, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 11798 + }, + { + "epoch": 9.4241214057508, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 11799 + }, + { + "epoch": 9.424920127795527, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 11800 + }, + { + "epoch": 9.425718849840255, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 11801 + }, + { + "epoch": 9.426517571884984, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 11802 + }, + { + "epoch": 9.427316293929712, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11803 + }, + { + "epoch": 9.428115015974441, + "grad_norm": 0.0927734375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 11804 + }, + { + "epoch": 9.428913738019169, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 11805 + }, + { + "epoch": 9.429712460063898, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 11806 + }, + { + "epoch": 9.430511182108626, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 11807 + }, + { + "epoch": 9.431309904153355, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 11808 + }, + { + "epoch": 9.432108626198083, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 11809 + }, + { + "epoch": 9.43290734824281, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 11810 + }, + { + "epoch": 9.43370607028754, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 11811 + }, + { + "epoch": 9.434504792332268, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11812 + }, + { + "epoch": 9.435303514376997, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 11813 + }, + { + "epoch": 9.436102236421725, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 11814 + }, + { + "epoch": 9.436900958466454, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 11815 + }, + { + "epoch": 9.437699680511182, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1155, + "step": 11816 + }, + { + "epoch": 9.438498402555911, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 11817 + }, + { + "epoch": 9.439297124600639, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 11818 + }, + { + "epoch": 9.440095846645367, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 11819 + }, + { + "epoch": 9.440894568690096, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 11820 + }, + { + "epoch": 9.441693290734824, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 11821 + }, + { + "epoch": 9.442492012779553, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 11822 + }, + { + "epoch": 9.44329073482428, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 11823 + }, + { + "epoch": 9.44408945686901, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 11824 + }, + { + "epoch": 9.444888178913738, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 11825 + }, + { + "epoch": 9.445686900958467, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 11826 + }, + { + "epoch": 9.446485623003195, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 11827 + }, + { + "epoch": 9.447284345047922, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11828 + }, + { + "epoch": 9.448083067092652, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 11829 + }, + { + "epoch": 9.44888178913738, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 11830 + }, + { + "epoch": 9.449680511182109, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 11831 + }, + { + "epoch": 9.450479233226837, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 11832 + }, + { + "epoch": 9.451277955271566, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 11833 + }, + { + "epoch": 9.452076677316294, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 11834 + }, + { + "epoch": 9.452875399361023, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 11835 + }, + { + "epoch": 9.45367412140575, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 11836 + }, + { + "epoch": 9.454472843450478, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 11837 + }, + { + "epoch": 9.455271565495208, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 11838 + }, + { + "epoch": 9.456070287539935, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 11839 + }, + { + "epoch": 9.456869009584665, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 11840 + }, + { + "epoch": 9.457667731629392, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 11841 + }, + { + "epoch": 9.458466453674122, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 11842 + }, + { + "epoch": 9.45926517571885, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 11843 + }, + { + "epoch": 9.460063897763579, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11844 + }, + { + "epoch": 9.460862619808307, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 11845 + }, + { + "epoch": 9.461661341853036, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 11846 + }, + { + "epoch": 9.462460063897764, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11847 + }, + { + "epoch": 9.463258785942491, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 11848 + }, + { + "epoch": 9.46405750798722, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 11849 + }, + { + "epoch": 9.464856230031948, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 11850 + }, + { + "epoch": 9.465654952076678, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 11851 + }, + { + "epoch": 9.466453674121405, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11852 + }, + { + "epoch": 9.467252396166135, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 11853 + }, + { + "epoch": 9.468051118210862, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 11854 + }, + { + "epoch": 9.468849840255592, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 11855 + }, + { + "epoch": 9.46964856230032, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 11856 + }, + { + "epoch": 9.470447284345047, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 11857 + }, + { + "epoch": 9.471246006389777, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 11858 + }, + { + "epoch": 9.472044728434504, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 11859 + }, + { + "epoch": 9.472843450479234, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 11860 + }, + { + "epoch": 9.473642172523961, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 11861 + }, + { + "epoch": 9.47444089456869, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 11862 + }, + { + "epoch": 9.475239616613418, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11863 + }, + { + "epoch": 9.476038338658148, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 11864 + }, + { + "epoch": 9.476837060702875, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 11865 + }, + { + "epoch": 9.477635782747603, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 11866 + }, + { + "epoch": 9.478434504792332, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 11867 + }, + { + "epoch": 9.47923322683706, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 11868 + }, + { + "epoch": 9.48003194888179, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 11869 + }, + { + "epoch": 9.480830670926517, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11870 + }, + { + "epoch": 9.481629392971247, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 11871 + }, + { + "epoch": 9.482428115015974, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 11872 + }, + { + "epoch": 9.483226837060704, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 11873 + }, + { + "epoch": 9.484025559105431, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 11874 + }, + { + "epoch": 9.48482428115016, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 11875 + }, + { + "epoch": 9.485623003194888, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 11876 + }, + { + "epoch": 9.486421725239616, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 11877 + }, + { + "epoch": 9.487220447284345, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 11878 + }, + { + "epoch": 9.488019169329073, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 11879 + }, + { + "epoch": 9.488817891373802, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 11880 + }, + { + "epoch": 9.48961661341853, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 11881 + }, + { + "epoch": 9.49041533546326, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 11882 + }, + { + "epoch": 9.491214057507987, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 11883 + }, + { + "epoch": 9.492012779552716, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 11884 + }, + { + "epoch": 9.492811501597444, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 11885 + }, + { + "epoch": 9.493610223642172, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 11886 + }, + { + "epoch": 9.494408945686901, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 11887 + }, + { + "epoch": 9.495207667731629, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 11888 + }, + { + "epoch": 9.496006389776358, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 11889 + }, + { + "epoch": 9.496805111821086, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 11890 + }, + { + "epoch": 9.497603833865815, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 11891 + }, + { + "epoch": 9.498402555910543, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 11892 + }, + { + "epoch": 9.499201277955272, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 11893 + }, + { + "epoch": 9.5, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 11894 + }, + { + "epoch": 9.500798722044728, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11895 + }, + { + "epoch": 9.501597444089457, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 11896 + }, + { + "epoch": 9.502396166134185, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 11897 + }, + { + "epoch": 9.503194888178914, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 11898 + }, + { + "epoch": 9.503993610223642, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11899 + }, + { + "epoch": 9.504792332268371, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 11900 + }, + { + "epoch": 9.505591054313099, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 11901 + }, + { + "epoch": 9.506389776357828, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11902 + }, + { + "epoch": 9.507188498402556, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 11903 + }, + { + "epoch": 9.507987220447284, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 11904 + }, + { + "epoch": 9.508785942492013, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 11905 + }, + { + "epoch": 9.50958466453674, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 11906 + }, + { + "epoch": 9.51038338658147, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 11907 + }, + { + "epoch": 9.511182108626198, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 11908 + }, + { + "epoch": 9.511980830670927, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 11909 + }, + { + "epoch": 9.512779552715655, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 11910 + }, + { + "epoch": 9.513578274760384, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 11911 + }, + { + "epoch": 9.514376996805112, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 11912 + }, + { + "epoch": 9.51517571884984, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 11913 + }, + { + "epoch": 9.515974440894569, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 11914 + }, + { + "epoch": 9.516773162939296, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 11915 + }, + { + "epoch": 9.517571884984026, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 11916 + }, + { + "epoch": 9.518370607028753, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 11917 + }, + { + "epoch": 9.519169329073483, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 11918 + }, + { + "epoch": 9.51996805111821, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 11919 + }, + { + "epoch": 9.52076677316294, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 11920 + }, + { + "epoch": 9.521565495207668, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 11921 + }, + { + "epoch": 9.522364217252395, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 11922 + }, + { + "epoch": 9.523162939297125, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 11923 + }, + { + "epoch": 9.523961661341852, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 11924 + }, + { + "epoch": 9.524760383386582, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 11925 + }, + { + "epoch": 9.52555910543131, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 11926 + }, + { + "epoch": 9.526357827476039, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 11927 + }, + { + "epoch": 9.527156549520766, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 11928 + }, + { + "epoch": 9.527955271565496, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 11929 + }, + { + "epoch": 9.528753993610223, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 11930 + }, + { + "epoch": 9.529552715654953, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 11931 + }, + { + "epoch": 9.53035143769968, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 11932 + }, + { + "epoch": 9.531150159744408, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 11933 + }, + { + "epoch": 9.531948881789138, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 11934 + }, + { + "epoch": 9.532747603833865, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 11935 + }, + { + "epoch": 9.533546325878595, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 11936 + }, + { + "epoch": 9.534345047923322, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 11937 + }, + { + "epoch": 9.535143769968052, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 11938 + }, + { + "epoch": 9.53594249201278, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 11939 + }, + { + "epoch": 9.536741214057509, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 11940 + }, + { + "epoch": 9.537539936102236, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 11941 + }, + { + "epoch": 9.538338658146966, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11942 + }, + { + "epoch": 9.539137380191693, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 11943 + }, + { + "epoch": 9.539936102236421, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 11944 + }, + { + "epoch": 9.54073482428115, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 11945 + }, + { + "epoch": 9.541533546325878, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 11946 + }, + { + "epoch": 9.542332268370608, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 11947 + }, + { + "epoch": 9.543130990415335, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 11948 + }, + { + "epoch": 9.543929712460065, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 11949 + }, + { + "epoch": 9.544728434504792, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 11950 + }, + { + "epoch": 9.545527156549522, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 11951 + }, + { + "epoch": 9.54632587859425, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 11952 + }, + { + "epoch": 9.547124600638977, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 11953 + }, + { + "epoch": 9.547923322683706, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 11954 + }, + { + "epoch": 9.548722044728434, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 11955 + }, + { + "epoch": 9.549520766773163, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 11956 + }, + { + "epoch": 9.550319488817891, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11957 + }, + { + "epoch": 9.55111821086262, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 11958 + }, + { + "epoch": 9.551916932907348, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 11959 + }, + { + "epoch": 9.552715654952078, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 11960 + }, + { + "epoch": 9.553514376996805, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 11961 + }, + { + "epoch": 9.554313099041533, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 11962 + }, + { + "epoch": 9.555111821086262, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 11963 + }, + { + "epoch": 9.55591054313099, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 11964 + }, + { + "epoch": 9.55670926517572, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 11965 + }, + { + "epoch": 9.557507987220447, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 11966 + }, + { + "epoch": 9.558306709265176, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 11967 + }, + { + "epoch": 9.559105431309904, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 11968 + }, + { + "epoch": 9.559904153354633, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 11969 + }, + { + "epoch": 9.560702875399361, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 11970 + }, + { + "epoch": 9.561501597444089, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 11971 + }, + { + "epoch": 9.562300319488818, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 11972 + }, + { + "epoch": 9.563099041533546, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 11973 + }, + { + "epoch": 9.563897763578275, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 11974 + }, + { + "epoch": 9.564696485623003, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 11975 + }, + { + "epoch": 9.565495207667732, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 11976 + }, + { + "epoch": 9.56629392971246, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 11977 + }, + { + "epoch": 9.56709265175719, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 11978 + }, + { + "epoch": 9.567891373801917, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 11979 + }, + { + "epoch": 9.568690095846645, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 11980 + }, + { + "epoch": 9.569488817891374, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 11981 + }, + { + "epoch": 9.570287539936102, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 11982 + }, + { + "epoch": 9.571086261980831, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 11983 + }, + { + "epoch": 9.571884984025559, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 11984 + }, + { + "epoch": 9.572683706070288, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 11985 + }, + { + "epoch": 9.573482428115016, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 11986 + }, + { + "epoch": 9.574281150159745, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 11987 + }, + { + "epoch": 9.575079872204473, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 11988 + }, + { + "epoch": 9.5758785942492, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 11989 + }, + { + "epoch": 9.57667731629393, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 11990 + }, + { + "epoch": 9.577476038338657, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 11991 + }, + { + "epoch": 9.578274760383387, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 11992 + }, + { + "epoch": 9.579073482428115, + "grad_norm": 0.09765625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 11993 + }, + { + "epoch": 9.579872204472844, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 11994 + }, + { + "epoch": 9.580670926517572, + "grad_norm": 0.09423828125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 11995 + }, + { + "epoch": 9.581469648562301, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 11996 + }, + { + "epoch": 9.582268370607029, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 11997 + }, + { + "epoch": 9.583067092651756, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 11998 + }, + { + "epoch": 9.583865814696486, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 11999 + }, + { + "epoch": 9.584664536741213, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 12000 + }, + { + "epoch": 9.585463258785943, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 12001 + }, + { + "epoch": 9.58626198083067, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 12002 + }, + { + "epoch": 9.5870607028754, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 12003 + }, + { + "epoch": 9.587859424920127, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 12004 + }, + { + "epoch": 9.588658146964857, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 12005 + }, + { + "epoch": 9.589456869009584, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 12006 + }, + { + "epoch": 9.590255591054314, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 12007 + }, + { + "epoch": 9.591054313099042, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 12008 + }, + { + "epoch": 9.59185303514377, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 12009 + }, + { + "epoch": 9.592651757188499, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 12010 + }, + { + "epoch": 9.593450479233226, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 12011 + }, + { + "epoch": 9.594249201277956, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 12012 + }, + { + "epoch": 9.595047923322683, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 12013 + }, + { + "epoch": 9.595846645367413, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 12014 + }, + { + "epoch": 9.59664536741214, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12015 + }, + { + "epoch": 9.59744408945687, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 12016 + }, + { + "epoch": 9.598242811501597, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 12017 + }, + { + "epoch": 9.599041533546325, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 12018 + }, + { + "epoch": 9.599840255591054, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12019 + }, + { + "epoch": 9.600638977635782, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 12020 + }, + { + "epoch": 9.601437699680512, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 12021 + }, + { + "epoch": 9.60223642172524, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 12022 + }, + { + "epoch": 9.603035143769969, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 12023 + }, + { + "epoch": 9.603833865814696, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 12024 + }, + { + "epoch": 9.604632587859426, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 12025 + }, + { + "epoch": 9.605431309904153, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 12026 + }, + { + "epoch": 9.606230031948883, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 12027 + }, + { + "epoch": 9.60702875399361, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 12028 + }, + { + "epoch": 9.607827476038338, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 12029 + }, + { + "epoch": 9.608626198083067, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 12030 + }, + { + "epoch": 9.609424920127795, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 12031 + }, + { + "epoch": 9.610223642172524, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 12032 + }, + { + "epoch": 9.611022364217252, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1125, + "step": 12033 + }, + { + "epoch": 9.611821086261982, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 12034 + }, + { + "epoch": 9.61261980830671, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 12035 + }, + { + "epoch": 9.613418530351439, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 12036 + }, + { + "epoch": 9.614217252396166, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 12037 + }, + { + "epoch": 9.615015974440894, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12038 + }, + { + "epoch": 9.615814696485623, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 12039 + }, + { + "epoch": 9.616613418530351, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 12040 + }, + { + "epoch": 9.61741214057508, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 12041 + }, + { + "epoch": 9.618210862619808, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 12042 + }, + { + "epoch": 9.619009584664537, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 12043 + }, + { + "epoch": 9.619808306709265, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 12044 + }, + { + "epoch": 9.620607028753994, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 12045 + }, + { + "epoch": 9.621405750798722, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 12046 + }, + { + "epoch": 9.62220447284345, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 12047 + }, + { + "epoch": 9.62300319488818, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 12048 + }, + { + "epoch": 9.623801916932907, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 12049 + }, + { + "epoch": 9.624600638977636, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 12050 + }, + { + "epoch": 9.625399361022364, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 12051 + }, + { + "epoch": 9.626198083067093, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 12052 + }, + { + "epoch": 9.62699680511182, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 12053 + }, + { + "epoch": 9.62779552715655, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 12054 + }, + { + "epoch": 9.628594249201278, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 12055 + }, + { + "epoch": 9.629392971246006, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 12056 + }, + { + "epoch": 9.630191693290735, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 12057 + }, + { + "epoch": 9.630990415335463, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 12058 + }, + { + "epoch": 9.631789137380192, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 12059 + }, + { + "epoch": 9.63258785942492, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 12060 + }, + { + "epoch": 9.633386581469649, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 12061 + }, + { + "epoch": 9.634185303514377, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 12062 + }, + { + "epoch": 9.634984025559106, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 12063 + }, + { + "epoch": 9.635782747603834, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 12064 + }, + { + "epoch": 9.636581469648561, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 12065 + }, + { + "epoch": 9.63738019169329, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 12066 + }, + { + "epoch": 9.638178913738018, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 12067 + }, + { + "epoch": 9.638977635782748, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 12068 + }, + { + "epoch": 9.639776357827476, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 12069 + }, + { + "epoch": 9.640575079872205, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 12070 + }, + { + "epoch": 9.641373801916933, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 12071 + }, + { + "epoch": 9.642172523961662, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1119, + "step": 12072 + }, + { + "epoch": 9.64297124600639, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 12073 + }, + { + "epoch": 9.643769968051117, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 12074 + }, + { + "epoch": 9.644568690095847, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 12075 + }, + { + "epoch": 9.645367412140574, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 12076 + }, + { + "epoch": 9.646166134185304, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 12077 + }, + { + "epoch": 9.646964856230031, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 12078 + }, + { + "epoch": 9.64776357827476, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 12079 + }, + { + "epoch": 9.648562300319488, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 12080 + }, + { + "epoch": 9.649361022364218, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 12081 + }, + { + "epoch": 9.650159744408946, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 12082 + }, + { + "epoch": 9.650958466453675, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 12083 + }, + { + "epoch": 9.651757188498403, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 12084 + }, + { + "epoch": 9.65255591054313, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 12085 + }, + { + "epoch": 9.65335463258786, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 12086 + }, + { + "epoch": 9.654153354632587, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 12087 + }, + { + "epoch": 9.654952076677317, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 12088 + }, + { + "epoch": 9.655750798722044, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 12089 + }, + { + "epoch": 9.656549520766774, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 12090 + }, + { + "epoch": 9.657348242811501, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 12091 + }, + { + "epoch": 9.65814696485623, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 12092 + }, + { + "epoch": 9.658945686900958, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1133, + "step": 12093 + }, + { + "epoch": 9.659744408945686, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 12094 + }, + { + "epoch": 9.660543130990416, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 12095 + }, + { + "epoch": 9.661341853035143, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 12096 + }, + { + "epoch": 9.662140575079873, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 12097 + }, + { + "epoch": 9.6629392971246, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 12098 + }, + { + "epoch": 9.66373801916933, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 12099 + }, + { + "epoch": 9.664536741214057, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 12100 + }, + { + "epoch": 9.665335463258787, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 12101 + }, + { + "epoch": 9.666134185303514, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 12102 + }, + { + "epoch": 9.666932907348244, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 12103 + }, + { + "epoch": 9.667731629392971, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 12104 + }, + { + "epoch": 9.668530351437699, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 12105 + }, + { + "epoch": 9.669329073482428, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 12106 + }, + { + "epoch": 9.670127795527156, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 12107 + }, + { + "epoch": 9.670926517571885, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 12108 + }, + { + "epoch": 9.671725239616613, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 12109 + }, + { + "epoch": 9.672523961661343, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 12110 + }, + { + "epoch": 9.67332268370607, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 12111 + }, + { + "epoch": 9.6741214057508, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 12112 + }, + { + "epoch": 9.674920127795527, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 12113 + }, + { + "epoch": 9.675718849840255, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 12114 + }, + { + "epoch": 9.676517571884984, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 12115 + }, + { + "epoch": 9.677316293929712, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 12116 + }, + { + "epoch": 9.678115015974441, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 12117 + }, + { + "epoch": 9.678913738019169, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12118 + }, + { + "epoch": 9.679712460063898, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12119 + }, + { + "epoch": 9.680511182108626, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 12120 + }, + { + "epoch": 9.681309904153355, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 12121 + }, + { + "epoch": 9.682108626198083, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 12122 + }, + { + "epoch": 9.68290734824281, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.115, + "step": 12123 + }, + { + "epoch": 9.68370607028754, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 12124 + }, + { + "epoch": 9.684504792332268, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 12125 + }, + { + "epoch": 9.685303514376997, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 12126 + }, + { + "epoch": 9.686102236421725, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 12127 + }, + { + "epoch": 9.686900958466454, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 12128 + }, + { + "epoch": 9.687699680511182, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 12129 + }, + { + "epoch": 9.688498402555911, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 12130 + }, + { + "epoch": 9.689297124600639, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 12131 + }, + { + "epoch": 9.690095846645367, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 12132 + }, + { + "epoch": 9.690894568690096, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 12133 + }, + { + "epoch": 9.691693290734824, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 12134 + }, + { + "epoch": 9.692492012779553, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 12135 + }, + { + "epoch": 9.69329073482428, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 12136 + }, + { + "epoch": 9.69408945686901, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 12137 + }, + { + "epoch": 9.694888178913738, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 12138 + }, + { + "epoch": 9.695686900958467, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 12139 + }, + { + "epoch": 9.696485623003195, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 12140 + }, + { + "epoch": 9.697284345047922, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 12141 + }, + { + "epoch": 9.698083067092652, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 12142 + }, + { + "epoch": 9.69888178913738, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 12143 + }, + { + "epoch": 9.699680511182109, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 12144 + }, + { + "epoch": 9.700479233226837, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 12145 + }, + { + "epoch": 9.701277955271566, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 12146 + }, + { + "epoch": 9.702076677316294, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12147 + }, + { + "epoch": 9.702875399361023, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 12148 + }, + { + "epoch": 9.70367412140575, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 12149 + }, + { + "epoch": 9.704472843450478, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 12150 + }, + { + "epoch": 9.705271565495208, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 12151 + }, + { + "epoch": 9.706070287539935, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 12152 + }, + { + "epoch": 9.706869009584665, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 12153 + }, + { + "epoch": 9.707667731629392, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 12154 + }, + { + "epoch": 9.708466453674122, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 12155 + }, + { + "epoch": 9.70926517571885, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 12156 + }, + { + "epoch": 9.710063897763579, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 12157 + }, + { + "epoch": 9.710862619808307, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12158 + }, + { + "epoch": 9.711661341853034, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 12159 + }, + { + "epoch": 9.712460063897764, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 12160 + }, + { + "epoch": 9.713258785942491, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 12161 + }, + { + "epoch": 9.71405750798722, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 12162 + }, + { + "epoch": 9.714856230031948, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 12163 + }, + { + "epoch": 9.715654952076678, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 12164 + }, + { + "epoch": 9.716453674121405, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 12165 + }, + { + "epoch": 9.717252396166135, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 12166 + }, + { + "epoch": 9.718051118210862, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 12167 + }, + { + "epoch": 9.718849840255592, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 12168 + }, + { + "epoch": 9.71964856230032, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 12169 + }, + { + "epoch": 9.720447284345047, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 12170 + }, + { + "epoch": 9.721246006389777, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 12171 + }, + { + "epoch": 9.722044728434504, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 12172 + }, + { + "epoch": 9.722843450479234, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12173 + }, + { + "epoch": 9.723642172523961, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 12174 + }, + { + "epoch": 9.72444089456869, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 12175 + }, + { + "epoch": 9.725239616613418, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 12176 + }, + { + "epoch": 9.726038338658148, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 12177 + }, + { + "epoch": 9.726837060702875, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 12178 + }, + { + "epoch": 9.727635782747605, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1115, + "step": 12179 + }, + { + "epoch": 9.728434504792332, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 12180 + }, + { + "epoch": 9.72923322683706, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12181 + }, + { + "epoch": 9.73003194888179, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 12182 + }, + { + "epoch": 9.730830670926517, + "grad_norm": 0.396484375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 12183 + }, + { + "epoch": 9.731629392971247, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 12184 + }, + { + "epoch": 9.732428115015974, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 12185 + }, + { + "epoch": 9.733226837060704, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 12186 + }, + { + "epoch": 9.734025559105431, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 12187 + }, + { + "epoch": 9.73482428115016, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 12188 + }, + { + "epoch": 9.735623003194888, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 12189 + }, + { + "epoch": 9.736421725239616, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 12190 + }, + { + "epoch": 9.737220447284345, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 12191 + }, + { + "epoch": 9.738019169329073, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 12192 + }, + { + "epoch": 9.738817891373802, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 12193 + }, + { + "epoch": 9.73961661341853, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 12194 + }, + { + "epoch": 9.74041533546326, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 12195 + }, + { + "epoch": 9.741214057507987, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 12196 + }, + { + "epoch": 9.742012779552716, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 12197 + }, + { + "epoch": 9.742811501597444, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 12198 + }, + { + "epoch": 9.743610223642172, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 12199 + }, + { + "epoch": 9.744408945686901, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 12200 + }, + { + "epoch": 9.745207667731629, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 12201 + }, + { + "epoch": 9.746006389776358, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 12202 + }, + { + "epoch": 9.746805111821086, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 12203 + }, + { + "epoch": 9.747603833865815, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 12204 + }, + { + "epoch": 9.748402555910543, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 12205 + }, + { + "epoch": 9.749201277955272, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 12206 + }, + { + "epoch": 9.75, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 12207 + }, + { + "epoch": 9.750798722044728, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 12208 + }, + { + "epoch": 9.751597444089457, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 12209 + }, + { + "epoch": 9.752396166134185, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 12210 + }, + { + "epoch": 9.753194888178914, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 12211 + }, + { + "epoch": 9.753993610223642, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 12212 + }, + { + "epoch": 9.754792332268371, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 12213 + }, + { + "epoch": 9.755591054313099, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 12214 + }, + { + "epoch": 9.756389776357828, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 12215 + }, + { + "epoch": 9.757188498402556, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 12216 + }, + { + "epoch": 9.757987220447284, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 12217 + }, + { + "epoch": 9.758785942492013, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 12218 + }, + { + "epoch": 9.75958466453674, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12219 + }, + { + "epoch": 9.76038338658147, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 12220 + }, + { + "epoch": 9.761182108626198, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 12221 + }, + { + "epoch": 9.761980830670927, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 12222 + }, + { + "epoch": 9.762779552715655, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 12223 + }, + { + "epoch": 9.763578274760384, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 12224 + }, + { + "epoch": 9.764376996805112, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 12225 + }, + { + "epoch": 9.76517571884984, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 12226 + }, + { + "epoch": 9.765974440894569, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 12227 + }, + { + "epoch": 9.766773162939296, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 12228 + }, + { + "epoch": 9.767571884984026, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 12229 + }, + { + "epoch": 9.768370607028753, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 12230 + }, + { + "epoch": 9.769169329073483, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 12231 + }, + { + "epoch": 9.76996805111821, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 12232 + }, + { + "epoch": 9.77076677316294, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 12233 + }, + { + "epoch": 9.771565495207668, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 12234 + }, + { + "epoch": 9.772364217252395, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 12235 + }, + { + "epoch": 9.773162939297125, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 12236 + }, + { + "epoch": 9.773961661341852, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 12237 + }, + { + "epoch": 9.774760383386582, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 12238 + }, + { + "epoch": 9.77555910543131, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 12239 + }, + { + "epoch": 9.776357827476039, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 12240 + }, + { + "epoch": 9.777156549520766, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 12241 + }, + { + "epoch": 9.777955271565496, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 12242 + }, + { + "epoch": 9.778753993610223, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 12243 + }, + { + "epoch": 9.779552715654953, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 12244 + }, + { + "epoch": 9.78035143769968, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 12245 + }, + { + "epoch": 9.781150159744408, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 12246 + }, + { + "epoch": 9.781948881789138, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 12247 + }, + { + "epoch": 9.782747603833865, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 12248 + }, + { + "epoch": 9.783546325878595, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 12249 + }, + { + "epoch": 9.784345047923322, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 12250 + }, + { + "epoch": 9.785143769968052, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 12251 + }, + { + "epoch": 9.78594249201278, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 12252 + }, + { + "epoch": 9.786741214057509, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 12253 + }, + { + "epoch": 9.787539936102236, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 12254 + }, + { + "epoch": 9.788338658146966, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 12255 + }, + { + "epoch": 9.789137380191693, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12256 + }, + { + "epoch": 9.789936102236421, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 12257 + }, + { + "epoch": 9.79073482428115, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 12258 + }, + { + "epoch": 9.791533546325878, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12259 + }, + { + "epoch": 9.792332268370608, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 12260 + }, + { + "epoch": 9.793130990415335, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 12261 + }, + { + "epoch": 9.793929712460065, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 12262 + }, + { + "epoch": 9.794728434504792, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 12263 + }, + { + "epoch": 9.795527156549522, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 12264 + }, + { + "epoch": 9.79632587859425, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 12265 + }, + { + "epoch": 9.797124600638977, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 12266 + }, + { + "epoch": 9.797923322683706, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1143, + "step": 12267 + }, + { + "epoch": 9.798722044728434, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 12268 + }, + { + "epoch": 9.799520766773163, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 12269 + }, + { + "epoch": 9.800319488817891, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 12270 + }, + { + "epoch": 9.80111821086262, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 12271 + }, + { + "epoch": 9.801916932907348, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 12272 + }, + { + "epoch": 9.802715654952078, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 12273 + }, + { + "epoch": 9.803514376996805, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 12274 + }, + { + "epoch": 9.804313099041533, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 12275 + }, + { + "epoch": 9.805111821086262, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 12276 + }, + { + "epoch": 9.80591054313099, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 12277 + }, + { + "epoch": 9.80670926517572, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 12278 + }, + { + "epoch": 9.807507987220447, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 12279 + }, + { + "epoch": 9.808306709265176, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 12280 + }, + { + "epoch": 9.809105431309904, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 12281 + }, + { + "epoch": 9.809904153354633, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 12282 + }, + { + "epoch": 9.810702875399361, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 12283 + }, + { + "epoch": 9.811501597444089, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 12284 + }, + { + "epoch": 9.812300319488818, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 12285 + }, + { + "epoch": 9.813099041533546, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 12286 + }, + { + "epoch": 9.813897763578275, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 12287 + }, + { + "epoch": 9.814696485623003, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 12288 + }, + { + "epoch": 9.815495207667732, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 12289 + }, + { + "epoch": 9.81629392971246, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 12290 + }, + { + "epoch": 9.81709265175719, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 12291 + }, + { + "epoch": 9.817891373801917, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 12292 + }, + { + "epoch": 9.818690095846645, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 12293 + }, + { + "epoch": 9.819488817891374, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 12294 + }, + { + "epoch": 9.820287539936102, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 12295 + }, + { + "epoch": 9.821086261980831, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 12296 + }, + { + "epoch": 9.821884984025559, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 12297 + }, + { + "epoch": 9.822683706070288, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 12298 + }, + { + "epoch": 9.823482428115016, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 12299 + }, + { + "epoch": 9.824281150159745, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 12300 + }, + { + "epoch": 9.825079872204473, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 12301 + }, + { + "epoch": 9.8258785942492, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12302 + }, + { + "epoch": 9.82667731629393, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 12303 + }, + { + "epoch": 9.827476038338657, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 12304 + }, + { + "epoch": 9.828274760383387, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 12305 + }, + { + "epoch": 9.829073482428115, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 12306 + }, + { + "epoch": 9.829872204472844, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 12307 + }, + { + "epoch": 9.830670926517572, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 12308 + }, + { + "epoch": 9.831469648562301, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 12309 + }, + { + "epoch": 9.832268370607029, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 12310 + }, + { + "epoch": 9.833067092651756, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 12311 + }, + { + "epoch": 9.833865814696486, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 12312 + }, + { + "epoch": 9.834664536741213, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 12313 + }, + { + "epoch": 9.835463258785943, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 12314 + }, + { + "epoch": 9.83626198083067, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 12315 + }, + { + "epoch": 9.8370607028754, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 12316 + }, + { + "epoch": 9.837859424920127, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 12317 + }, + { + "epoch": 9.838658146964857, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 12318 + }, + { + "epoch": 9.839456869009584, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 12319 + }, + { + "epoch": 9.840255591054314, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 12320 + }, + { + "epoch": 9.841054313099042, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 12321 + }, + { + "epoch": 9.84185303514377, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 12322 + }, + { + "epoch": 9.842651757188499, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 12323 + }, + { + "epoch": 9.843450479233226, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 12324 + }, + { + "epoch": 9.844249201277956, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 12325 + }, + { + "epoch": 9.845047923322683, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 12326 + }, + { + "epoch": 9.845846645367413, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 12327 + }, + { + "epoch": 9.84664536741214, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 12328 + }, + { + "epoch": 9.84744408945687, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 12329 + }, + { + "epoch": 9.848242811501597, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 12330 + }, + { + "epoch": 9.849041533546325, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 12331 + }, + { + "epoch": 9.849840255591054, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 12332 + }, + { + "epoch": 9.850638977635782, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 12333 + }, + { + "epoch": 9.851437699680512, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 12334 + }, + { + "epoch": 9.85223642172524, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 12335 + }, + { + "epoch": 9.853035143769969, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 12336 + }, + { + "epoch": 9.853833865814696, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12337 + }, + { + "epoch": 9.854632587859426, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 12338 + }, + { + "epoch": 9.855431309904153, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 12339 + }, + { + "epoch": 9.856230031948883, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 12340 + }, + { + "epoch": 9.85702875399361, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 12341 + }, + { + "epoch": 9.857827476038338, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 12342 + }, + { + "epoch": 9.858626198083067, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 12343 + }, + { + "epoch": 9.859424920127795, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 12344 + }, + { + "epoch": 9.860223642172524, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 12345 + }, + { + "epoch": 9.861022364217252, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 12346 + }, + { + "epoch": 9.861821086261982, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 12347 + }, + { + "epoch": 9.86261980830671, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 12348 + }, + { + "epoch": 9.863418530351439, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 12349 + }, + { + "epoch": 9.864217252396166, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12350 + }, + { + "epoch": 9.865015974440894, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 12351 + }, + { + "epoch": 9.865814696485623, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 12352 + }, + { + "epoch": 9.866613418530351, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 12353 + }, + { + "epoch": 9.86741214057508, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 12354 + }, + { + "epoch": 9.868210862619808, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 12355 + }, + { + "epoch": 9.869009584664537, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 12356 + }, + { + "epoch": 9.869808306709265, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 12357 + }, + { + "epoch": 9.870607028753994, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 12358 + }, + { + "epoch": 9.871405750798722, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 12359 + }, + { + "epoch": 9.87220447284345, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 12360 + }, + { + "epoch": 9.87300319488818, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 12361 + }, + { + "epoch": 9.873801916932907, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 12362 + }, + { + "epoch": 9.874600638977636, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 12363 + }, + { + "epoch": 9.875399361022364, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 12364 + }, + { + "epoch": 9.876198083067093, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 12365 + }, + { + "epoch": 9.87699680511182, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 12366 + }, + { + "epoch": 9.87779552715655, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 12367 + }, + { + "epoch": 9.878594249201278, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 12368 + }, + { + "epoch": 9.879392971246006, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12369 + }, + { + "epoch": 9.880191693290735, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 12370 + }, + { + "epoch": 9.880990415335463, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 12371 + }, + { + "epoch": 9.881789137380192, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 12372 + }, + { + "epoch": 9.88258785942492, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 12373 + }, + { + "epoch": 9.883386581469649, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 12374 + }, + { + "epoch": 9.884185303514377, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 12375 + }, + { + "epoch": 9.884984025559106, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 12376 + }, + { + "epoch": 9.885782747603834, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12377 + }, + { + "epoch": 9.886581469648561, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 12378 + }, + { + "epoch": 9.88738019169329, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 12379 + }, + { + "epoch": 9.888178913738018, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 12380 + }, + { + "epoch": 9.888977635782748, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 12381 + }, + { + "epoch": 9.889776357827476, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 12382 + }, + { + "epoch": 9.890575079872205, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 12383 + }, + { + "epoch": 9.891373801916933, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 12384 + }, + { + "epoch": 9.892172523961662, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 12385 + }, + { + "epoch": 9.89297124600639, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 12386 + }, + { + "epoch": 9.893769968051117, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 12387 + }, + { + "epoch": 9.894568690095847, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 12388 + }, + { + "epoch": 9.895367412140574, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 12389 + }, + { + "epoch": 9.896166134185304, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 12390 + }, + { + "epoch": 9.896964856230031, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 12391 + }, + { + "epoch": 9.89776357827476, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 12392 + }, + { + "epoch": 9.898562300319488, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 12393 + }, + { + "epoch": 9.899361022364218, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 12394 + }, + { + "epoch": 9.900159744408946, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 12395 + }, + { + "epoch": 9.900958466453675, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 12396 + }, + { + "epoch": 9.901757188498403, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 12397 + }, + { + "epoch": 9.90255591054313, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12398 + }, + { + "epoch": 9.90335463258786, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 12399 + }, + { + "epoch": 9.904153354632587, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 12400 + }, + { + "epoch": 9.904952076677317, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 12401 + }, + { + "epoch": 9.905750798722044, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 12402 + }, + { + "epoch": 9.906549520766774, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12403 + }, + { + "epoch": 9.907348242811501, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 12404 + }, + { + "epoch": 9.90814696485623, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 12405 + }, + { + "epoch": 9.908945686900958, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 12406 + }, + { + "epoch": 9.909744408945686, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 12407 + }, + { + "epoch": 9.910543130990416, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 12408 + }, + { + "epoch": 9.911341853035143, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 12409 + }, + { + "epoch": 9.912140575079873, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 12410 + }, + { + "epoch": 9.9129392971246, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 12411 + }, + { + "epoch": 9.91373801916933, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 12412 + }, + { + "epoch": 9.914536741214057, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 12413 + }, + { + "epoch": 9.915335463258787, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 12414 + }, + { + "epoch": 9.916134185303514, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 12415 + }, + { + "epoch": 9.916932907348244, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 12416 + }, + { + "epoch": 9.917731629392971, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 12417 + }, + { + "epoch": 9.918530351437699, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 12418 + }, + { + "epoch": 9.919329073482428, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 12419 + }, + { + "epoch": 9.920127795527156, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 12420 + }, + { + "epoch": 9.920926517571885, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 12421 + }, + { + "epoch": 9.921725239616613, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 12422 + }, + { + "epoch": 9.922523961661343, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 12423 + }, + { + "epoch": 9.92332268370607, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 12424 + }, + { + "epoch": 9.9241214057508, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 12425 + }, + { + "epoch": 9.924920127795527, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 12426 + }, + { + "epoch": 9.925718849840255, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 12427 + }, + { + "epoch": 9.926517571884984, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 12428 + }, + { + "epoch": 9.927316293929712, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 12429 + }, + { + "epoch": 9.928115015974441, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 12430 + }, + { + "epoch": 9.928913738019169, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 12431 + }, + { + "epoch": 9.929712460063898, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 12432 + }, + { + "epoch": 9.930511182108626, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 12433 + }, + { + "epoch": 9.931309904153355, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 12434 + }, + { + "epoch": 9.932108626198083, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 12435 + }, + { + "epoch": 9.93290734824281, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 12436 + }, + { + "epoch": 9.93370607028754, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 12437 + }, + { + "epoch": 9.934504792332268, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12438 + }, + { + "epoch": 9.935303514376997, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 12439 + }, + { + "epoch": 9.936102236421725, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 12440 + }, + { + "epoch": 9.936900958466454, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 12441 + }, + { + "epoch": 9.937699680511182, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 12442 + }, + { + "epoch": 9.938498402555911, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 12443 + }, + { + "epoch": 9.939297124600639, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 12444 + }, + { + "epoch": 9.940095846645367, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 12445 + }, + { + "epoch": 9.940894568690096, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12446 + }, + { + "epoch": 9.941693290734824, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 12447 + }, + { + "epoch": 9.942492012779553, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 12448 + }, + { + "epoch": 9.94329073482428, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 12449 + }, + { + "epoch": 9.94408945686901, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 12450 + }, + { + "epoch": 9.944888178913738, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 12451 + }, + { + "epoch": 9.945686900958467, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 12452 + }, + { + "epoch": 9.946485623003195, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 12453 + }, + { + "epoch": 9.947284345047922, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 12454 + }, + { + "epoch": 9.948083067092652, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 12455 + }, + { + "epoch": 9.94888178913738, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 12456 + }, + { + "epoch": 9.949680511182109, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12457 + }, + { + "epoch": 9.950479233226837, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 12458 + }, + { + "epoch": 9.951277955271566, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 12459 + }, + { + "epoch": 9.952076677316294, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 12460 + }, + { + "epoch": 9.952875399361023, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 12461 + }, + { + "epoch": 9.95367412140575, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 12462 + }, + { + "epoch": 9.954472843450478, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 12463 + }, + { + "epoch": 9.955271565495208, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 12464 + }, + { + "epoch": 9.956070287539935, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 12465 + }, + { + "epoch": 9.956869009584665, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 12466 + }, + { + "epoch": 9.957667731629392, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 12467 + }, + { + "epoch": 9.958466453674122, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 12468 + }, + { + "epoch": 9.95926517571885, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 12469 + }, + { + "epoch": 9.960063897763579, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 12470 + }, + { + "epoch": 9.960862619808307, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 12471 + }, + { + "epoch": 9.961661341853034, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 12472 + }, + { + "epoch": 9.962460063897764, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 12473 + }, + { + "epoch": 9.963258785942491, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 12474 + }, + { + "epoch": 9.96405750798722, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 12475 + }, + { + "epoch": 9.964856230031948, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 12476 + }, + { + "epoch": 9.965654952076678, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 12477 + }, + { + "epoch": 9.966453674121405, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 12478 + }, + { + "epoch": 9.967252396166135, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 12479 + }, + { + "epoch": 9.968051118210862, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 12480 + }, + { + "epoch": 9.968849840255592, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 12481 + }, + { + "epoch": 9.96964856230032, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 12482 + }, + { + "epoch": 9.970447284345047, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 12483 + }, + { + "epoch": 9.971246006389777, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 12484 + }, + { + "epoch": 9.972044728434504, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 12485 + }, + { + "epoch": 9.972843450479234, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12486 + }, + { + "epoch": 9.973642172523961, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 12487 + }, + { + "epoch": 9.97444089456869, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 12488 + }, + { + "epoch": 9.975239616613418, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 12489 + }, + { + "epoch": 9.976038338658148, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 12490 + }, + { + "epoch": 9.976837060702875, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 12491 + }, + { + "epoch": 9.977635782747605, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 12492 + }, + { + "epoch": 9.978434504792332, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 12493 + }, + { + "epoch": 9.97923322683706, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 12494 + }, + { + "epoch": 9.98003194888179, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12495 + }, + { + "epoch": 9.980830670926517, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 12496 + }, + { + "epoch": 9.981629392971247, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 12497 + }, + { + "epoch": 9.982428115015974, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 12498 + }, + { + "epoch": 9.983226837060704, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 12499 + }, + { + "epoch": 9.984025559105431, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 12500 + }, + { + "epoch": 9.98482428115016, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 12501 + }, + { + "epoch": 9.985623003194888, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 12502 + }, + { + "epoch": 9.986421725239616, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 12503 + }, + { + "epoch": 9.987220447284345, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 12504 + }, + { + "epoch": 9.988019169329073, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 12505 + }, + { + "epoch": 9.988817891373802, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 12506 + }, + { + "epoch": 9.98961661341853, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 12507 + }, + { + "epoch": 9.99041533546326, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 12508 + }, + { + "epoch": 9.991214057507987, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 12509 + }, + { + "epoch": 9.992012779552716, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 12510 + }, + { + "epoch": 9.992811501597444, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 12511 + }, + { + "epoch": 9.993610223642172, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 12512 + }, + { + "epoch": 9.994408945686901, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 12513 + }, + { + "epoch": 9.995207667731629, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 12514 + }, + { + "epoch": 9.996006389776358, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 12515 + }, + { + "epoch": 9.996805111821086, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 12516 + }, + { + "epoch": 9.997603833865815, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 12517 + }, + { + "epoch": 9.998402555910543, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 12518 + }, + { + "epoch": 9.999201277955272, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 12519 + }, + { + "epoch": 10.0, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 12520 + }, + { + "epoch": 10.000798722044728, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 12521 + }, + { + "epoch": 10.001597444089457, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 12522 + }, + { + "epoch": 10.002396166134185, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 12523 + }, + { + "epoch": 10.003194888178914, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 12524 + }, + { + "epoch": 10.003993610223642, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 12525 + }, + { + "epoch": 10.004792332268371, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12526 + }, + { + "epoch": 10.005591054313099, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 12527 + }, + { + "epoch": 10.006389776357828, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 12528 + }, + { + "epoch": 10.007188498402556, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 12529 + }, + { + "epoch": 10.007987220447284, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 12530 + }, + { + "epoch": 10.008785942492013, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 12531 + }, + { + "epoch": 10.00958466453674, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 12532 + }, + { + "epoch": 10.01038338658147, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 12533 + }, + { + "epoch": 10.011182108626198, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 12534 + }, + { + "epoch": 10.011980830670927, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 12535 + }, + { + "epoch": 10.012779552715655, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 12536 + }, + { + "epoch": 10.013578274760384, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 12537 + }, + { + "epoch": 10.014376996805112, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 12538 + }, + { + "epoch": 10.01517571884984, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 12539 + }, + { + "epoch": 10.015974440894569, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 12540 + }, + { + "epoch": 10.016773162939296, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 12541 + }, + { + "epoch": 10.017571884984026, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 12542 + }, + { + "epoch": 10.018370607028753, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 12543 + }, + { + "epoch": 10.019169329073483, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 12544 + }, + { + "epoch": 10.01996805111821, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 12545 + }, + { + "epoch": 10.02076677316294, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12546 + }, + { + "epoch": 10.021565495207668, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12547 + }, + { + "epoch": 10.022364217252397, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 12548 + }, + { + "epoch": 10.023162939297125, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 12549 + }, + { + "epoch": 10.023961661341852, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 12550 + }, + { + "epoch": 10.024760383386582, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 12551 + }, + { + "epoch": 10.02555910543131, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 12552 + }, + { + "epoch": 10.026357827476039, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 12553 + }, + { + "epoch": 10.027156549520766, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 12554 + }, + { + "epoch": 10.027955271565496, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 12555 + }, + { + "epoch": 10.028753993610223, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 12556 + }, + { + "epoch": 10.029552715654953, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 12557 + }, + { + "epoch": 10.03035143769968, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12558 + }, + { + "epoch": 10.031150159744408, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 12559 + }, + { + "epoch": 10.031948881789138, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12560 + }, + { + "epoch": 10.032747603833865, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 12561 + }, + { + "epoch": 10.033546325878595, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 12562 + }, + { + "epoch": 10.034345047923322, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 12563 + }, + { + "epoch": 10.035143769968052, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 12564 + }, + { + "epoch": 10.03594249201278, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 12565 + }, + { + "epoch": 10.036741214057509, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 12566 + }, + { + "epoch": 10.037539936102236, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 12567 + }, + { + "epoch": 10.038338658146964, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 12568 + }, + { + "epoch": 10.039137380191693, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 12569 + }, + { + "epoch": 10.039936102236421, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 12570 + }, + { + "epoch": 10.04073482428115, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 12571 + }, + { + "epoch": 10.041533546325878, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 12572 + }, + { + "epoch": 10.042332268370608, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 12573 + }, + { + "epoch": 10.043130990415335, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 12574 + }, + { + "epoch": 10.043929712460065, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 12575 + }, + { + "epoch": 10.044728434504792, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 12576 + }, + { + "epoch": 10.04552715654952, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 12577 + }, + { + "epoch": 10.04632587859425, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 12578 + }, + { + "epoch": 10.047124600638977, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 12579 + }, + { + "epoch": 10.047923322683706, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 12580 + }, + { + "epoch": 10.048722044728434, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 12581 + }, + { + "epoch": 10.049520766773163, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12582 + }, + { + "epoch": 10.050319488817891, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 12583 + }, + { + "epoch": 10.05111821086262, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 12584 + }, + { + "epoch": 10.051916932907348, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 12585 + }, + { + "epoch": 10.052715654952078, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 12586 + }, + { + "epoch": 10.053514376996805, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 12587 + }, + { + "epoch": 10.054313099041533, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 12588 + }, + { + "epoch": 10.055111821086262, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 12589 + }, + { + "epoch": 10.05591054313099, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1203, + "step": 12590 + }, + { + "epoch": 10.05670926517572, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 12591 + }, + { + "epoch": 10.057507987220447, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 12592 + }, + { + "epoch": 10.058306709265176, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 12593 + }, + { + "epoch": 10.059105431309904, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 12594 + }, + { + "epoch": 10.059904153354633, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 12595 + }, + { + "epoch": 10.060702875399361, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 12596 + }, + { + "epoch": 10.061501597444089, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 12597 + }, + { + "epoch": 10.062300319488818, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12598 + }, + { + "epoch": 10.063099041533546, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 12599 + }, + { + "epoch": 10.063897763578275, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 12600 + }, + { + "epoch": 10.064696485623003, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 12601 + }, + { + "epoch": 10.065495207667732, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12602 + }, + { + "epoch": 10.06629392971246, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12603 + }, + { + "epoch": 10.06709265175719, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 12604 + }, + { + "epoch": 10.067891373801917, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 12605 + }, + { + "epoch": 10.068690095846645, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 12606 + }, + { + "epoch": 10.069488817891374, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 12607 + }, + { + "epoch": 10.070287539936102, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 12608 + }, + { + "epoch": 10.071086261980831, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 12609 + }, + { + "epoch": 10.071884984025559, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 12610 + }, + { + "epoch": 10.072683706070288, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 12611 + }, + { + "epoch": 10.073482428115016, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 12612 + }, + { + "epoch": 10.074281150159745, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1175, + "step": 12613 + }, + { + "epoch": 10.075079872204473, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 12614 + }, + { + "epoch": 10.0758785942492, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 12615 + }, + { + "epoch": 10.07667731629393, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 12616 + }, + { + "epoch": 10.077476038338657, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 12617 + }, + { + "epoch": 10.078274760383387, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 12618 + }, + { + "epoch": 10.079073482428115, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 12619 + }, + { + "epoch": 10.079872204472844, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 12620 + }, + { + "epoch": 10.080670926517572, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 12621 + }, + { + "epoch": 10.081469648562301, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 12622 + }, + { + "epoch": 10.082268370607029, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 12623 + }, + { + "epoch": 10.083067092651758, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 12624 + }, + { + "epoch": 10.083865814696486, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 12625 + }, + { + "epoch": 10.084664536741213, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 12626 + }, + { + "epoch": 10.085463258785943, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 12627 + }, + { + "epoch": 10.08626198083067, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 12628 + }, + { + "epoch": 10.0870607028754, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 12629 + }, + { + "epoch": 10.087859424920127, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 12630 + }, + { + "epoch": 10.088658146964857, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 12631 + }, + { + "epoch": 10.089456869009584, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 12632 + }, + { + "epoch": 10.090255591054314, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12633 + }, + { + "epoch": 10.091054313099042, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 12634 + }, + { + "epoch": 10.09185303514377, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 12635 + }, + { + "epoch": 10.092651757188499, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 12636 + }, + { + "epoch": 10.093450479233226, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 12637 + }, + { + "epoch": 10.094249201277956, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 12638 + }, + { + "epoch": 10.095047923322683, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 12639 + }, + { + "epoch": 10.095846645367413, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 12640 + }, + { + "epoch": 10.09664536741214, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 12641 + }, + { + "epoch": 10.09744408945687, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 12642 + }, + { + "epoch": 10.098242811501597, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 12643 + }, + { + "epoch": 10.099041533546325, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 12644 + }, + { + "epoch": 10.099840255591054, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 12645 + }, + { + "epoch": 10.100638977635782, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 12646 + }, + { + "epoch": 10.101437699680512, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 12647 + }, + { + "epoch": 10.10223642172524, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 12648 + }, + { + "epoch": 10.103035143769969, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 12649 + }, + { + "epoch": 10.103833865814696, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1197, + "step": 12650 + }, + { + "epoch": 10.104632587859426, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 12651 + }, + { + "epoch": 10.105431309904153, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 12652 + }, + { + "epoch": 10.106230031948881, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 12653 + }, + { + "epoch": 10.10702875399361, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 12654 + }, + { + "epoch": 10.107827476038338, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 12655 + }, + { + "epoch": 10.108626198083067, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 12656 + }, + { + "epoch": 10.109424920127795, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 12657 + }, + { + "epoch": 10.110223642172524, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 12658 + }, + { + "epoch": 10.111022364217252, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 12659 + }, + { + "epoch": 10.111821086261982, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 12660 + }, + { + "epoch": 10.11261980830671, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 12661 + }, + { + "epoch": 10.113418530351439, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 12662 + }, + { + "epoch": 10.114217252396166, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 12663 + }, + { + "epoch": 10.115015974440894, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 12664 + }, + { + "epoch": 10.115814696485623, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 12665 + }, + { + "epoch": 10.116613418530351, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 12666 + }, + { + "epoch": 10.11741214057508, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 12667 + }, + { + "epoch": 10.118210862619808, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 12668 + }, + { + "epoch": 10.119009584664537, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 12669 + }, + { + "epoch": 10.119808306709265, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 12670 + }, + { + "epoch": 10.120607028753994, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 12671 + }, + { + "epoch": 10.121405750798722, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 12672 + }, + { + "epoch": 10.12220447284345, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 12673 + }, + { + "epoch": 10.12300319488818, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 12674 + }, + { + "epoch": 10.123801916932907, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 12675 + }, + { + "epoch": 10.124600638977636, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 12676 + }, + { + "epoch": 10.125399361022364, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 12677 + }, + { + "epoch": 10.126198083067093, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 12678 + }, + { + "epoch": 10.12699680511182, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 12679 + }, + { + "epoch": 10.12779552715655, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 12680 + }, + { + "epoch": 10.128594249201278, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 12681 + }, + { + "epoch": 10.129392971246006, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 12682 + }, + { + "epoch": 10.130191693290735, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 12683 + }, + { + "epoch": 10.130990415335463, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 12684 + }, + { + "epoch": 10.131789137380192, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 12685 + }, + { + "epoch": 10.13258785942492, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 12686 + }, + { + "epoch": 10.133386581469649, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 12687 + }, + { + "epoch": 10.134185303514377, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 12688 + }, + { + "epoch": 10.134984025559106, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 12689 + }, + { + "epoch": 10.135782747603834, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 12690 + }, + { + "epoch": 10.136581469648561, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12691 + }, + { + "epoch": 10.13738019169329, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 12692 + }, + { + "epoch": 10.138178913738018, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 12693 + }, + { + "epoch": 10.138977635782748, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 12694 + }, + { + "epoch": 10.139776357827476, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 12695 + }, + { + "epoch": 10.140575079872205, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 12696 + }, + { + "epoch": 10.141373801916933, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 12697 + }, + { + "epoch": 10.142172523961662, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 12698 + }, + { + "epoch": 10.14297124600639, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 12699 + }, + { + "epoch": 10.143769968051119, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12700 + }, + { + "epoch": 10.144568690095847, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 12701 + }, + { + "epoch": 10.145367412140574, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 12702 + }, + { + "epoch": 10.146166134185304, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 12703 + }, + { + "epoch": 10.146964856230031, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 12704 + }, + { + "epoch": 10.14776357827476, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 12705 + }, + { + "epoch": 10.148562300319488, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 12706 + }, + { + "epoch": 10.149361022364218, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 12707 + }, + { + "epoch": 10.150159744408946, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 12708 + }, + { + "epoch": 10.150958466453675, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 12709 + }, + { + "epoch": 10.151757188498403, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 12710 + }, + { + "epoch": 10.15255591054313, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 12711 + }, + { + "epoch": 10.15335463258786, + "grad_norm": 0.09375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 12712 + }, + { + "epoch": 10.154153354632587, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 12713 + }, + { + "epoch": 10.154952076677317, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 12714 + }, + { + "epoch": 10.155750798722044, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 12715 + }, + { + "epoch": 10.156549520766774, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 12716 + }, + { + "epoch": 10.157348242811501, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 12717 + }, + { + "epoch": 10.15814696485623, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 12718 + }, + { + "epoch": 10.158945686900958, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 12719 + }, + { + "epoch": 10.159744408945686, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 12720 + }, + { + "epoch": 10.160543130990416, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 12721 + }, + { + "epoch": 10.161341853035143, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 12722 + }, + { + "epoch": 10.162140575079873, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 12723 + }, + { + "epoch": 10.1629392971246, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 12724 + }, + { + "epoch": 10.16373801916933, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 12725 + }, + { + "epoch": 10.164536741214057, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 12726 + }, + { + "epoch": 10.165335463258787, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 12727 + }, + { + "epoch": 10.166134185303514, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12728 + }, + { + "epoch": 10.166932907348242, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 12729 + }, + { + "epoch": 10.167731629392971, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 12730 + }, + { + "epoch": 10.168530351437699, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 12731 + }, + { + "epoch": 10.169329073482428, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 12732 + }, + { + "epoch": 10.170127795527156, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12733 + }, + { + "epoch": 10.170926517571885, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 12734 + }, + { + "epoch": 10.171725239616613, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 12735 + }, + { + "epoch": 10.172523961661343, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 12736 + }, + { + "epoch": 10.17332268370607, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 12737 + }, + { + "epoch": 10.1741214057508, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12738 + }, + { + "epoch": 10.174920127795527, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 12739 + }, + { + "epoch": 10.175718849840255, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 12740 + }, + { + "epoch": 10.176517571884984, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 12741 + }, + { + "epoch": 10.177316293929712, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 12742 + }, + { + "epoch": 10.178115015974441, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 12743 + }, + { + "epoch": 10.178913738019169, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 12744 + }, + { + "epoch": 10.179712460063898, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 12745 + }, + { + "epoch": 10.180511182108626, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 12746 + }, + { + "epoch": 10.181309904153355, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 12747 + }, + { + "epoch": 10.182108626198083, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 12748 + }, + { + "epoch": 10.18290734824281, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 12749 + }, + { + "epoch": 10.18370607028754, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.1139, + "step": 12750 + }, + { + "epoch": 10.184504792332268, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 12751 + }, + { + "epoch": 10.185303514376997, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 12752 + }, + { + "epoch": 10.186102236421725, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 12753 + }, + { + "epoch": 10.186900958466454, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 12754 + }, + { + "epoch": 10.187699680511182, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 12755 + }, + { + "epoch": 10.188498402555911, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 12756 + }, + { + "epoch": 10.189297124600639, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 12757 + }, + { + "epoch": 10.190095846645367, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 12758 + }, + { + "epoch": 10.190894568690096, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 12759 + }, + { + "epoch": 10.191693290734824, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 12760 + }, + { + "epoch": 10.192492012779553, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 12761 + }, + { + "epoch": 10.19329073482428, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 12762 + }, + { + "epoch": 10.19408945686901, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 12763 + }, + { + "epoch": 10.194888178913738, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 12764 + }, + { + "epoch": 10.195686900958467, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 12765 + }, + { + "epoch": 10.196485623003195, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 12766 + }, + { + "epoch": 10.197284345047922, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 12767 + }, + { + "epoch": 10.198083067092652, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 12768 + }, + { + "epoch": 10.19888178913738, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 12769 + }, + { + "epoch": 10.199680511182109, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 12770 + }, + { + "epoch": 10.200479233226837, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 12771 + }, + { + "epoch": 10.201277955271566, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 12772 + }, + { + "epoch": 10.202076677316294, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 12773 + }, + { + "epoch": 10.202875399361023, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 12774 + }, + { + "epoch": 10.20367412140575, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 12775 + }, + { + "epoch": 10.204472843450478, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 12776 + }, + { + "epoch": 10.205271565495208, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 12777 + }, + { + "epoch": 10.206070287539935, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 12778 + }, + { + "epoch": 10.206869009584665, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 12779 + }, + { + "epoch": 10.207667731629392, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 12780 + }, + { + "epoch": 10.208466453674122, + "grad_norm": 0.095703125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 12781 + }, + { + "epoch": 10.20926517571885, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 12782 + }, + { + "epoch": 10.210063897763579, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 12783 + }, + { + "epoch": 10.210862619808307, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 12784 + }, + { + "epoch": 10.211661341853036, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 12785 + }, + { + "epoch": 10.212460063897764, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 12786 + }, + { + "epoch": 10.213258785942491, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 12787 + }, + { + "epoch": 10.21405750798722, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 12788 + }, + { + "epoch": 10.214856230031948, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 12789 + }, + { + "epoch": 10.215654952076678, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12790 + }, + { + "epoch": 10.216453674121405, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 12791 + }, + { + "epoch": 10.217252396166135, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 12792 + }, + { + "epoch": 10.218051118210862, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 12793 + }, + { + "epoch": 10.218849840255592, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 12794 + }, + { + "epoch": 10.21964856230032, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 12795 + }, + { + "epoch": 10.220447284345047, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 12796 + }, + { + "epoch": 10.221246006389777, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 12797 + }, + { + "epoch": 10.222044728434504, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 12798 + }, + { + "epoch": 10.222843450479234, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 12799 + }, + { + "epoch": 10.223642172523961, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 12800 + }, + { + "epoch": 10.22444089456869, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 12801 + }, + { + "epoch": 10.225239616613418, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 12802 + }, + { + "epoch": 10.226038338658148, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 12803 + }, + { + "epoch": 10.226837060702875, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 12804 + }, + { + "epoch": 10.227635782747603, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 12805 + }, + { + "epoch": 10.228434504792332, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 12806 + }, + { + "epoch": 10.22923322683706, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 12807 + }, + { + "epoch": 10.23003194888179, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 12808 + }, + { + "epoch": 10.230830670926517, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 12809 + }, + { + "epoch": 10.231629392971247, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 12810 + }, + { + "epoch": 10.232428115015974, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 12811 + }, + { + "epoch": 10.233226837060704, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 12812 + }, + { + "epoch": 10.234025559105431, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 12813 + }, + { + "epoch": 10.23482428115016, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 12814 + }, + { + "epoch": 10.235623003194888, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 12815 + }, + { + "epoch": 10.236421725239616, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 12816 + }, + { + "epoch": 10.237220447284345, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 12817 + }, + { + "epoch": 10.238019169329073, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 12818 + }, + { + "epoch": 10.238817891373802, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 12819 + }, + { + "epoch": 10.23961661341853, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 12820 + }, + { + "epoch": 10.24041533546326, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 12821 + }, + { + "epoch": 10.241214057507987, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 12822 + }, + { + "epoch": 10.242012779552716, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 12823 + }, + { + "epoch": 10.242811501597444, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 12824 + }, + { + "epoch": 10.243610223642172, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 12825 + }, + { + "epoch": 10.244408945686901, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 12826 + }, + { + "epoch": 10.245207667731629, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 12827 + }, + { + "epoch": 10.246006389776358, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 12828 + }, + { + "epoch": 10.246805111821086, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 12829 + }, + { + "epoch": 10.247603833865815, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 12830 + }, + { + "epoch": 10.248402555910543, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 12831 + }, + { + "epoch": 10.249201277955272, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 12832 + }, + { + "epoch": 10.25, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 12833 + }, + { + "epoch": 10.250798722044728, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 12834 + }, + { + "epoch": 10.251597444089457, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 12835 + }, + { + "epoch": 10.252396166134185, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12836 + }, + { + "epoch": 10.253194888178914, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 12837 + }, + { + "epoch": 10.253993610223642, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 12838 + }, + { + "epoch": 10.254792332268371, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 12839 + }, + { + "epoch": 10.255591054313099, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 12840 + }, + { + "epoch": 10.256389776357828, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 12841 + }, + { + "epoch": 10.257188498402556, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 12842 + }, + { + "epoch": 10.257987220447284, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 12843 + }, + { + "epoch": 10.258785942492013, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 12844 + }, + { + "epoch": 10.25958466453674, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 12845 + }, + { + "epoch": 10.26038338658147, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12846 + }, + { + "epoch": 10.261182108626198, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 12847 + }, + { + "epoch": 10.261980830670927, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 12848 + }, + { + "epoch": 10.262779552715655, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 12849 + }, + { + "epoch": 10.263578274760384, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 12850 + }, + { + "epoch": 10.264376996805112, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 12851 + }, + { + "epoch": 10.26517571884984, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 12852 + }, + { + "epoch": 10.265974440894569, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 12853 + }, + { + "epoch": 10.266773162939296, + "grad_norm": 0.095703125, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 12854 + }, + { + "epoch": 10.267571884984026, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 12855 + }, + { + "epoch": 10.268370607028753, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 12856 + }, + { + "epoch": 10.269169329073483, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 12857 + }, + { + "epoch": 10.26996805111821, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 12858 + }, + { + "epoch": 10.27076677316294, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 12859 + }, + { + "epoch": 10.271565495207668, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 12860 + }, + { + "epoch": 10.272364217252397, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 12861 + }, + { + "epoch": 10.273162939297125, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 12862 + }, + { + "epoch": 10.273961661341852, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 12863 + }, + { + "epoch": 10.274760383386582, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 12864 + }, + { + "epoch": 10.27555910543131, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 12865 + }, + { + "epoch": 10.276357827476039, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 12866 + }, + { + "epoch": 10.277156549520766, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 12867 + }, + { + "epoch": 10.277955271565496, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 12868 + }, + { + "epoch": 10.278753993610223, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 12869 + }, + { + "epoch": 10.279552715654953, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 12870 + }, + { + "epoch": 10.28035143769968, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 12871 + }, + { + "epoch": 10.281150159744408, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 12872 + }, + { + "epoch": 10.281948881789138, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 12873 + }, + { + "epoch": 10.282747603833865, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 12874 + }, + { + "epoch": 10.283546325878595, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 12875 + }, + { + "epoch": 10.284345047923322, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 12876 + }, + { + "epoch": 10.285143769968052, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 12877 + }, + { + "epoch": 10.28594249201278, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 12878 + }, + { + "epoch": 10.286741214057509, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 12879 + }, + { + "epoch": 10.287539936102236, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 12880 + }, + { + "epoch": 10.288338658146964, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 12881 + }, + { + "epoch": 10.289137380191693, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 12882 + }, + { + "epoch": 10.289936102236421, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 12883 + }, + { + "epoch": 10.29073482428115, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 12884 + }, + { + "epoch": 10.291533546325878, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 12885 + }, + { + "epoch": 10.292332268370608, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 12886 + }, + { + "epoch": 10.293130990415335, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 12887 + }, + { + "epoch": 10.293929712460065, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 12888 + }, + { + "epoch": 10.294728434504792, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 12889 + }, + { + "epoch": 10.295527156549522, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 12890 + }, + { + "epoch": 10.29632587859425, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12891 + }, + { + "epoch": 10.297124600638977, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 12892 + }, + { + "epoch": 10.297923322683706, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 12893 + }, + { + "epoch": 10.298722044728434, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 12894 + }, + { + "epoch": 10.299520766773163, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 12895 + }, + { + "epoch": 10.300319488817891, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 12896 + }, + { + "epoch": 10.30111821086262, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 12897 + }, + { + "epoch": 10.301916932907348, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 12898 + }, + { + "epoch": 10.302715654952078, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 12899 + }, + { + "epoch": 10.303514376996805, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 12900 + }, + { + "epoch": 10.304313099041533, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 12901 + }, + { + "epoch": 10.305111821086262, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 12902 + }, + { + "epoch": 10.30591054313099, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 12903 + }, + { + "epoch": 10.30670926517572, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 12904 + }, + { + "epoch": 10.307507987220447, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 12905 + }, + { + "epoch": 10.308306709265176, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 12906 + }, + { + "epoch": 10.309105431309904, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12907 + }, + { + "epoch": 10.309904153354633, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 12908 + }, + { + "epoch": 10.310702875399361, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 12909 + }, + { + "epoch": 10.311501597444089, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 12910 + }, + { + "epoch": 10.312300319488818, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 12911 + }, + { + "epoch": 10.313099041533546, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 12912 + }, + { + "epoch": 10.313897763578275, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 12913 + }, + { + "epoch": 10.314696485623003, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 12914 + }, + { + "epoch": 10.315495207667732, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 12915 + }, + { + "epoch": 10.31629392971246, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 12916 + }, + { + "epoch": 10.31709265175719, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 12917 + }, + { + "epoch": 10.317891373801917, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 12918 + }, + { + "epoch": 10.318690095846645, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 12919 + }, + { + "epoch": 10.319488817891374, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 12920 + }, + { + "epoch": 10.320287539936102, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 12921 + }, + { + "epoch": 10.321086261980831, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 12922 + }, + { + "epoch": 10.321884984025559, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 12923 + }, + { + "epoch": 10.322683706070288, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 12924 + }, + { + "epoch": 10.323482428115016, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 12925 + }, + { + "epoch": 10.324281150159745, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 12926 + }, + { + "epoch": 10.325079872204473, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 12927 + }, + { + "epoch": 10.3258785942492, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 12928 + }, + { + "epoch": 10.32667731629393, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 12929 + }, + { + "epoch": 10.327476038338657, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 12930 + }, + { + "epoch": 10.328274760383387, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 12931 + }, + { + "epoch": 10.329073482428115, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 12932 + }, + { + "epoch": 10.329872204472844, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 12933 + }, + { + "epoch": 10.330670926517572, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12934 + }, + { + "epoch": 10.331469648562301, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 12935 + }, + { + "epoch": 10.332268370607029, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 12936 + }, + { + "epoch": 10.333067092651758, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 12937 + }, + { + "epoch": 10.333865814696486, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 12938 + }, + { + "epoch": 10.334664536741213, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 12939 + }, + { + "epoch": 10.335463258785943, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 12940 + }, + { + "epoch": 10.33626198083067, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 12941 + }, + { + "epoch": 10.3370607028754, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 12942 + }, + { + "epoch": 10.337859424920127, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 12943 + }, + { + "epoch": 10.338658146964857, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 12944 + }, + { + "epoch": 10.339456869009584, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 12945 + }, + { + "epoch": 10.340255591054314, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 12946 + }, + { + "epoch": 10.341054313099042, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 12947 + }, + { + "epoch": 10.34185303514377, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 12948 + }, + { + "epoch": 10.342651757188499, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 12949 + }, + { + "epoch": 10.343450479233226, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12950 + }, + { + "epoch": 10.344249201277956, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 12951 + }, + { + "epoch": 10.345047923322683, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 12952 + }, + { + "epoch": 10.345846645367413, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 12953 + }, + { + "epoch": 10.34664536741214, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 12954 + }, + { + "epoch": 10.34744408945687, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 12955 + }, + { + "epoch": 10.348242811501597, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 12956 + }, + { + "epoch": 10.349041533546325, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 12957 + }, + { + "epoch": 10.349840255591054, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 12958 + }, + { + "epoch": 10.350638977635782, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 12959 + }, + { + "epoch": 10.351437699680512, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 12960 + }, + { + "epoch": 10.35223642172524, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 12961 + }, + { + "epoch": 10.353035143769969, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 12962 + }, + { + "epoch": 10.353833865814696, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 12963 + }, + { + "epoch": 10.354632587859426, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 12964 + }, + { + "epoch": 10.355431309904153, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 12965 + }, + { + "epoch": 10.356230031948883, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 12966 + }, + { + "epoch": 10.35702875399361, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 12967 + }, + { + "epoch": 10.357827476038338, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 12968 + }, + { + "epoch": 10.358626198083067, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 12969 + }, + { + "epoch": 10.359424920127795, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 12970 + }, + { + "epoch": 10.360223642172524, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 12971 + }, + { + "epoch": 10.361022364217252, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 12972 + }, + { + "epoch": 10.361821086261982, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 12973 + }, + { + "epoch": 10.36261980830671, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 12974 + }, + { + "epoch": 10.363418530351439, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 12975 + }, + { + "epoch": 10.364217252396166, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 12976 + }, + { + "epoch": 10.365015974440894, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 12977 + }, + { + "epoch": 10.365814696485623, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 12978 + }, + { + "epoch": 10.366613418530351, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 12979 + }, + { + "epoch": 10.36741214057508, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 12980 + }, + { + "epoch": 10.368210862619808, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 12981 + }, + { + "epoch": 10.369009584664537, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 12982 + }, + { + "epoch": 10.369808306709265, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 12983 + }, + { + "epoch": 10.370607028753994, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 12984 + }, + { + "epoch": 10.371405750798722, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 12985 + }, + { + "epoch": 10.37220447284345, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 12986 + }, + { + "epoch": 10.37300319488818, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 12987 + }, + { + "epoch": 10.373801916932907, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 12988 + }, + { + "epoch": 10.374600638977636, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 12989 + }, + { + "epoch": 10.375399361022364, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 12990 + }, + { + "epoch": 10.376198083067093, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 12991 + }, + { + "epoch": 10.37699680511182, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 12992 + }, + { + "epoch": 10.37779552715655, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 12993 + }, + { + "epoch": 10.378594249201278, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 12994 + }, + { + "epoch": 10.379392971246006, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 12995 + }, + { + "epoch": 10.380191693290735, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 12996 + }, + { + "epoch": 10.380990415335463, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 12997 + }, + { + "epoch": 10.381789137380192, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 12998 + }, + { + "epoch": 10.38258785942492, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 12999 + }, + { + "epoch": 10.383386581469649, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 13000 + }, + { + "epoch": 10.384185303514377, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 13001 + }, + { + "epoch": 10.384984025559106, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 13002 + }, + { + "epoch": 10.385782747603834, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 13003 + }, + { + "epoch": 10.386581469648561, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 13004 + }, + { + "epoch": 10.38738019169329, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 13005 + }, + { + "epoch": 10.388178913738018, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 13006 + }, + { + "epoch": 10.388977635782748, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 13007 + }, + { + "epoch": 10.389776357827476, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 13008 + }, + { + "epoch": 10.390575079872205, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 13009 + }, + { + "epoch": 10.391373801916933, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 13010 + }, + { + "epoch": 10.392172523961662, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 13011 + }, + { + "epoch": 10.39297124600639, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 13012 + }, + { + "epoch": 10.393769968051117, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13013 + }, + { + "epoch": 10.394568690095847, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 13014 + }, + { + "epoch": 10.395367412140574, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 13015 + }, + { + "epoch": 10.396166134185304, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 13016 + }, + { + "epoch": 10.396964856230031, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 13017 + }, + { + "epoch": 10.39776357827476, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 13018 + }, + { + "epoch": 10.398562300319488, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 13019 + }, + { + "epoch": 10.399361022364218, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 13020 + }, + { + "epoch": 10.400159744408946, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 13021 + }, + { + "epoch": 10.400958466453675, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 13022 + }, + { + "epoch": 10.401757188498403, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 13023 + }, + { + "epoch": 10.40255591054313, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 13024 + }, + { + "epoch": 10.40335463258786, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 13025 + }, + { + "epoch": 10.404153354632587, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 13026 + }, + { + "epoch": 10.404952076677317, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 13027 + }, + { + "epoch": 10.405750798722044, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 13028 + }, + { + "epoch": 10.406549520766774, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 13029 + }, + { + "epoch": 10.407348242811501, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 13030 + }, + { + "epoch": 10.40814696485623, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13031 + }, + { + "epoch": 10.408945686900958, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 13032 + }, + { + "epoch": 10.409744408945686, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 13033 + }, + { + "epoch": 10.410543130990416, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 13034 + }, + { + "epoch": 10.411341853035143, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 13035 + }, + { + "epoch": 10.412140575079873, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 13036 + }, + { + "epoch": 10.4129392971246, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 13037 + }, + { + "epoch": 10.41373801916933, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 13038 + }, + { + "epoch": 10.414536741214057, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 13039 + }, + { + "epoch": 10.415335463258787, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 13040 + }, + { + "epoch": 10.416134185303514, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 13041 + }, + { + "epoch": 10.416932907348242, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 13042 + }, + { + "epoch": 10.417731629392971, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 13043 + }, + { + "epoch": 10.418530351437699, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 13044 + }, + { + "epoch": 10.419329073482428, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 13045 + }, + { + "epoch": 10.420127795527156, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 13046 + }, + { + "epoch": 10.420926517571885, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 13047 + }, + { + "epoch": 10.421725239616613, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 13048 + }, + { + "epoch": 10.422523961661343, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 13049 + }, + { + "epoch": 10.42332268370607, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 13050 + }, + { + "epoch": 10.4241214057508, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13051 + }, + { + "epoch": 10.424920127795527, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 13052 + }, + { + "epoch": 10.425718849840255, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 13053 + }, + { + "epoch": 10.426517571884984, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 13054 + }, + { + "epoch": 10.427316293929712, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 13055 + }, + { + "epoch": 10.428115015974441, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 13056 + }, + { + "epoch": 10.428913738019169, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 13057 + }, + { + "epoch": 10.429712460063898, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 13058 + }, + { + "epoch": 10.430511182108626, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 13059 + }, + { + "epoch": 10.431309904153355, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 13060 + }, + { + "epoch": 10.432108626198083, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 13061 + }, + { + "epoch": 10.43290734824281, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 13062 + }, + { + "epoch": 10.43370607028754, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 13063 + }, + { + "epoch": 10.434504792332268, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 13064 + }, + { + "epoch": 10.435303514376997, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 13065 + }, + { + "epoch": 10.436102236421725, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 13066 + }, + { + "epoch": 10.436900958466454, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 13067 + }, + { + "epoch": 10.437699680511182, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 13068 + }, + { + "epoch": 10.438498402555911, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 13069 + }, + { + "epoch": 10.439297124600639, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 13070 + }, + { + "epoch": 10.440095846645367, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 13071 + }, + { + "epoch": 10.440894568690096, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 13072 + }, + { + "epoch": 10.441693290734824, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 13073 + }, + { + "epoch": 10.442492012779553, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 13074 + }, + { + "epoch": 10.44329073482428, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 13075 + }, + { + "epoch": 10.44408945686901, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 13076 + }, + { + "epoch": 10.444888178913738, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 13077 + }, + { + "epoch": 10.445686900958467, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 13078 + }, + { + "epoch": 10.446485623003195, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 13079 + }, + { + "epoch": 10.447284345047922, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13080 + }, + { + "epoch": 10.448083067092652, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 13081 + }, + { + "epoch": 10.44888178913738, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 13082 + }, + { + "epoch": 10.449680511182109, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 13083 + }, + { + "epoch": 10.450479233226837, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13084 + }, + { + "epoch": 10.451277955271566, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 13085 + }, + { + "epoch": 10.452076677316294, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 13086 + }, + { + "epoch": 10.452875399361023, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 13087 + }, + { + "epoch": 10.45367412140575, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 13088 + }, + { + "epoch": 10.454472843450478, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 13089 + }, + { + "epoch": 10.455271565495208, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1158, + "step": 13090 + }, + { + "epoch": 10.456070287539935, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 13091 + }, + { + "epoch": 10.456869009584665, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 13092 + }, + { + "epoch": 10.457667731629392, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 13093 + }, + { + "epoch": 10.458466453674122, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 13094 + }, + { + "epoch": 10.45926517571885, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 13095 + }, + { + "epoch": 10.460063897763579, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 13096 + }, + { + "epoch": 10.460862619808307, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 13097 + }, + { + "epoch": 10.461661341853036, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 13098 + }, + { + "epoch": 10.462460063897764, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 13099 + }, + { + "epoch": 10.463258785942491, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 13100 + }, + { + "epoch": 10.46405750798722, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 13101 + }, + { + "epoch": 10.464856230031948, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 13102 + }, + { + "epoch": 10.465654952076678, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 13103 + }, + { + "epoch": 10.466453674121405, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 13104 + }, + { + "epoch": 10.467252396166135, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 13105 + }, + { + "epoch": 10.468051118210862, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 13106 + }, + { + "epoch": 10.468849840255592, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 13107 + }, + { + "epoch": 10.46964856230032, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 13108 + }, + { + "epoch": 10.470447284345047, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 13109 + }, + { + "epoch": 10.471246006389777, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 13110 + }, + { + "epoch": 10.472044728434504, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13111 + }, + { + "epoch": 10.472843450479234, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 13112 + }, + { + "epoch": 10.473642172523961, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 13113 + }, + { + "epoch": 10.47444089456869, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 13114 + }, + { + "epoch": 10.475239616613418, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 13115 + }, + { + "epoch": 10.476038338658148, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 13116 + }, + { + "epoch": 10.476837060702875, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 13117 + }, + { + "epoch": 10.477635782747603, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 13118 + }, + { + "epoch": 10.478434504792332, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 13119 + }, + { + "epoch": 10.47923322683706, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 13120 + }, + { + "epoch": 10.48003194888179, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 13121 + }, + { + "epoch": 10.480830670926517, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 13122 + }, + { + "epoch": 10.481629392971247, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 13123 + }, + { + "epoch": 10.482428115015974, + "grad_norm": 0.09765625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 13124 + }, + { + "epoch": 10.483226837060704, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 13125 + }, + { + "epoch": 10.484025559105431, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 13126 + }, + { + "epoch": 10.48482428115016, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 13127 + }, + { + "epoch": 10.485623003194888, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 13128 + }, + { + "epoch": 10.486421725239616, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 13129 + }, + { + "epoch": 10.487220447284345, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 13130 + }, + { + "epoch": 10.488019169329073, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 13131 + }, + { + "epoch": 10.488817891373802, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 13132 + }, + { + "epoch": 10.48961661341853, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 13133 + }, + { + "epoch": 10.49041533546326, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 13134 + }, + { + "epoch": 10.491214057507987, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 13135 + }, + { + "epoch": 10.492012779552716, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 13136 + }, + { + "epoch": 10.492811501597444, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 13137 + }, + { + "epoch": 10.493610223642172, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 13138 + }, + { + "epoch": 10.494408945686901, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 13139 + }, + { + "epoch": 10.495207667731629, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 13140 + }, + { + "epoch": 10.496006389776358, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 13141 + }, + { + "epoch": 10.496805111821086, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 13142 + }, + { + "epoch": 10.497603833865815, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 13143 + }, + { + "epoch": 10.498402555910543, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 13144 + }, + { + "epoch": 10.499201277955272, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13145 + }, + { + "epoch": 10.5, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 13146 + }, + { + "epoch": 10.500798722044728, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 13147 + }, + { + "epoch": 10.501597444089457, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 13148 + }, + { + "epoch": 10.502396166134185, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 13149 + }, + { + "epoch": 10.503194888178914, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 13150 + }, + { + "epoch": 10.503993610223642, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 13151 + }, + { + "epoch": 10.504792332268371, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 13152 + }, + { + "epoch": 10.505591054313099, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 13153 + }, + { + "epoch": 10.506389776357828, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 13154 + }, + { + "epoch": 10.507188498402556, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 13155 + }, + { + "epoch": 10.507987220447284, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13156 + }, + { + "epoch": 10.508785942492013, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 13157 + }, + { + "epoch": 10.50958466453674, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 13158 + }, + { + "epoch": 10.51038338658147, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 13159 + }, + { + "epoch": 10.511182108626198, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 13160 + }, + { + "epoch": 10.511980830670927, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 13161 + }, + { + "epoch": 10.512779552715655, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 13162 + }, + { + "epoch": 10.513578274760384, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 13163 + }, + { + "epoch": 10.514376996805112, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 13164 + }, + { + "epoch": 10.51517571884984, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 13165 + }, + { + "epoch": 10.515974440894569, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 13166 + }, + { + "epoch": 10.516773162939296, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 13167 + }, + { + "epoch": 10.517571884984026, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 13168 + }, + { + "epoch": 10.518370607028753, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 13169 + }, + { + "epoch": 10.519169329073483, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 13170 + }, + { + "epoch": 10.51996805111821, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 13171 + }, + { + "epoch": 10.52076677316294, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 13172 + }, + { + "epoch": 10.521565495207668, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 13173 + }, + { + "epoch": 10.522364217252395, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 13174 + }, + { + "epoch": 10.523162939297125, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 13175 + }, + { + "epoch": 10.523961661341852, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 13176 + }, + { + "epoch": 10.524760383386582, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 13177 + }, + { + "epoch": 10.52555910543131, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13178 + }, + { + "epoch": 10.526357827476039, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 13179 + }, + { + "epoch": 10.527156549520766, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 13180 + }, + { + "epoch": 10.527955271565496, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 13181 + }, + { + "epoch": 10.528753993610223, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 13182 + }, + { + "epoch": 10.529552715654953, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 13183 + }, + { + "epoch": 10.53035143769968, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 13184 + }, + { + "epoch": 10.531150159744408, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 13185 + }, + { + "epoch": 10.531948881789138, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 13186 + }, + { + "epoch": 10.532747603833865, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 13187 + }, + { + "epoch": 10.533546325878595, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 13188 + }, + { + "epoch": 10.534345047923322, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 13189 + }, + { + "epoch": 10.535143769968052, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 13190 + }, + { + "epoch": 10.53594249201278, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 13191 + }, + { + "epoch": 10.536741214057509, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 13192 + }, + { + "epoch": 10.537539936102236, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 13193 + }, + { + "epoch": 10.538338658146966, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 13194 + }, + { + "epoch": 10.539137380191693, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 13195 + }, + { + "epoch": 10.539936102236421, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13196 + }, + { + "epoch": 10.54073482428115, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 13197 + }, + { + "epoch": 10.541533546325878, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 13198 + }, + { + "epoch": 10.542332268370608, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 13199 + }, + { + "epoch": 10.543130990415335, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 13200 + }, + { + "epoch": 10.543929712460065, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13201 + }, + { + "epoch": 10.544728434504792, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 13202 + }, + { + "epoch": 10.545527156549522, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 13203 + }, + { + "epoch": 10.54632587859425, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 13204 + }, + { + "epoch": 10.547124600638977, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 13205 + }, + { + "epoch": 10.547923322683706, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 13206 + }, + { + "epoch": 10.548722044728434, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 13207 + }, + { + "epoch": 10.549520766773163, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 13208 + }, + { + "epoch": 10.550319488817891, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 13209 + }, + { + "epoch": 10.55111821086262, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 13210 + }, + { + "epoch": 10.551916932907348, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 13211 + }, + { + "epoch": 10.552715654952078, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 13212 + }, + { + "epoch": 10.553514376996805, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 13213 + }, + { + "epoch": 10.554313099041533, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 13214 + }, + { + "epoch": 10.555111821086262, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 13215 + }, + { + "epoch": 10.55591054313099, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 13216 + }, + { + "epoch": 10.55670926517572, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 13217 + }, + { + "epoch": 10.557507987220447, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 13218 + }, + { + "epoch": 10.558306709265176, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 13219 + }, + { + "epoch": 10.559105431309904, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 13220 + }, + { + "epoch": 10.559904153354633, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 13221 + }, + { + "epoch": 10.560702875399361, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 13222 + }, + { + "epoch": 10.561501597444089, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 13223 + }, + { + "epoch": 10.562300319488818, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 13224 + }, + { + "epoch": 10.563099041533546, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 13225 + }, + { + "epoch": 10.563897763578275, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 13226 + }, + { + "epoch": 10.564696485623003, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 13227 + }, + { + "epoch": 10.565495207667732, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 13228 + }, + { + "epoch": 10.56629392971246, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 13229 + }, + { + "epoch": 10.56709265175719, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 13230 + }, + { + "epoch": 10.567891373801917, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 13231 + }, + { + "epoch": 10.568690095846645, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 13232 + }, + { + "epoch": 10.569488817891374, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13233 + }, + { + "epoch": 10.570287539936102, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 13234 + }, + { + "epoch": 10.571086261980831, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 13235 + }, + { + "epoch": 10.571884984025559, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 13236 + }, + { + "epoch": 10.572683706070288, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 13237 + }, + { + "epoch": 10.573482428115016, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 13238 + }, + { + "epoch": 10.574281150159745, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 13239 + }, + { + "epoch": 10.575079872204473, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 13240 + }, + { + "epoch": 10.5758785942492, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 13241 + }, + { + "epoch": 10.57667731629393, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 13242 + }, + { + "epoch": 10.577476038338657, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 13243 + }, + { + "epoch": 10.578274760383387, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 13244 + }, + { + "epoch": 10.579073482428115, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 13245 + }, + { + "epoch": 10.579872204472844, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 13246 + }, + { + "epoch": 10.580670926517572, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13247 + }, + { + "epoch": 10.581469648562301, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 13248 + }, + { + "epoch": 10.582268370607029, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 13249 + }, + { + "epoch": 10.583067092651756, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 13250 + }, + { + "epoch": 10.583865814696486, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 13251 + }, + { + "epoch": 10.584664536741213, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 13252 + }, + { + "epoch": 10.585463258785943, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 13253 + }, + { + "epoch": 10.58626198083067, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 13254 + }, + { + "epoch": 10.5870607028754, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 13255 + }, + { + "epoch": 10.587859424920127, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 13256 + }, + { + "epoch": 10.588658146964857, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 13257 + }, + { + "epoch": 10.589456869009584, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 13258 + }, + { + "epoch": 10.590255591054314, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 13259 + }, + { + "epoch": 10.591054313099042, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 13260 + }, + { + "epoch": 10.59185303514377, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 13261 + }, + { + "epoch": 10.592651757188499, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 13262 + }, + { + "epoch": 10.593450479233226, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 13263 + }, + { + "epoch": 10.594249201277956, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 13264 + }, + { + "epoch": 10.595047923322683, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 13265 + }, + { + "epoch": 10.595846645367413, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 13266 + }, + { + "epoch": 10.59664536741214, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 13267 + }, + { + "epoch": 10.59744408945687, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 13268 + }, + { + "epoch": 10.598242811501597, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 13269 + }, + { + "epoch": 10.599041533546325, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 13270 + }, + { + "epoch": 10.599840255591054, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 13271 + }, + { + "epoch": 10.600638977635782, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 13272 + }, + { + "epoch": 10.601437699680512, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 13273 + }, + { + "epoch": 10.60223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 13274 + }, + { + "epoch": 10.603035143769969, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 13275 + }, + { + "epoch": 10.603833865814696, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 13276 + }, + { + "epoch": 10.604632587859426, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 13277 + }, + { + "epoch": 10.605431309904153, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 13278 + }, + { + "epoch": 10.606230031948883, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 13279 + }, + { + "epoch": 10.60702875399361, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 13280 + }, + { + "epoch": 10.607827476038338, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 13281 + }, + { + "epoch": 10.608626198083067, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 13282 + }, + { + "epoch": 10.609424920127795, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13283 + }, + { + "epoch": 10.610223642172524, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 13284 + }, + { + "epoch": 10.611022364217252, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 13285 + }, + { + "epoch": 10.611821086261982, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 13286 + }, + { + "epoch": 10.61261980830671, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 13287 + }, + { + "epoch": 10.613418530351439, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 13288 + }, + { + "epoch": 10.614217252396166, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 13289 + }, + { + "epoch": 10.615015974440894, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 13290 + }, + { + "epoch": 10.615814696485623, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 13291 + }, + { + "epoch": 10.616613418530351, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 13292 + }, + { + "epoch": 10.61741214057508, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 13293 + }, + { + "epoch": 10.618210862619808, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 13294 + }, + { + "epoch": 10.619009584664537, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 13295 + }, + { + "epoch": 10.619808306709265, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 13296 + }, + { + "epoch": 10.620607028753994, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 13297 + }, + { + "epoch": 10.621405750798722, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 13298 + }, + { + "epoch": 10.62220447284345, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 13299 + }, + { + "epoch": 10.62300319488818, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 13300 + }, + { + "epoch": 10.623801916932907, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 13301 + }, + { + "epoch": 10.624600638977636, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 13302 + }, + { + "epoch": 10.625399361022364, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 13303 + }, + { + "epoch": 10.626198083067093, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 13304 + }, + { + "epoch": 10.62699680511182, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 13305 + }, + { + "epoch": 10.62779552715655, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 13306 + }, + { + "epoch": 10.628594249201278, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 13307 + }, + { + "epoch": 10.629392971246006, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 13308 + }, + { + "epoch": 10.630191693290735, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 13309 + }, + { + "epoch": 10.630990415335463, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 13310 + }, + { + "epoch": 10.631789137380192, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 13311 + }, + { + "epoch": 10.63258785942492, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 13312 + }, + { + "epoch": 10.633386581469649, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 13313 + }, + { + "epoch": 10.634185303514377, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 13314 + }, + { + "epoch": 10.634984025559106, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 13315 + }, + { + "epoch": 10.635782747603834, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 13316 + }, + { + "epoch": 10.636581469648561, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 13317 + }, + { + "epoch": 10.63738019169329, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 13318 + }, + { + "epoch": 10.638178913738018, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 13319 + }, + { + "epoch": 10.638977635782748, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 13320 + }, + { + "epoch": 10.639776357827476, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 13321 + }, + { + "epoch": 10.640575079872205, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 13322 + }, + { + "epoch": 10.641373801916933, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 13323 + }, + { + "epoch": 10.642172523961662, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 13324 + }, + { + "epoch": 10.64297124600639, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 13325 + }, + { + "epoch": 10.643769968051117, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 13326 + }, + { + "epoch": 10.644568690095847, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 13327 + }, + { + "epoch": 10.645367412140574, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 13328 + }, + { + "epoch": 10.646166134185304, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 13329 + }, + { + "epoch": 10.646964856230031, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 13330 + }, + { + "epoch": 10.64776357827476, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 13331 + }, + { + "epoch": 10.648562300319488, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13332 + }, + { + "epoch": 10.649361022364218, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 13333 + }, + { + "epoch": 10.650159744408946, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 13334 + }, + { + "epoch": 10.650958466453675, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 13335 + }, + { + "epoch": 10.651757188498403, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 13336 + }, + { + "epoch": 10.65255591054313, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 13337 + }, + { + "epoch": 10.65335463258786, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 13338 + }, + { + "epoch": 10.654153354632587, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 13339 + }, + { + "epoch": 10.654952076677317, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1145, + "step": 13340 + }, + { + "epoch": 10.655750798722044, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 13341 + }, + { + "epoch": 10.656549520766774, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 13342 + }, + { + "epoch": 10.657348242811501, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 13343 + }, + { + "epoch": 10.65814696485623, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 13344 + }, + { + "epoch": 10.658945686900958, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 13345 + }, + { + "epoch": 10.659744408945686, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 13346 + }, + { + "epoch": 10.660543130990416, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 13347 + }, + { + "epoch": 10.661341853035143, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 13348 + }, + { + "epoch": 10.662140575079873, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 13349 + }, + { + "epoch": 10.6629392971246, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 13350 + }, + { + "epoch": 10.66373801916933, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 13351 + }, + { + "epoch": 10.664536741214057, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 13352 + }, + { + "epoch": 10.665335463258787, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 13353 + }, + { + "epoch": 10.666134185303514, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 13354 + }, + { + "epoch": 10.666932907348244, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 13355 + }, + { + "epoch": 10.667731629392971, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 13356 + }, + { + "epoch": 10.668530351437699, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 13357 + }, + { + "epoch": 10.669329073482428, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 13358 + }, + { + "epoch": 10.670127795527156, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 13359 + }, + { + "epoch": 10.670926517571885, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 13360 + }, + { + "epoch": 10.671725239616613, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 13361 + }, + { + "epoch": 10.672523961661343, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 13362 + }, + { + "epoch": 10.67332268370607, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 13363 + }, + { + "epoch": 10.6741214057508, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 13364 + }, + { + "epoch": 10.674920127795527, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 13365 + }, + { + "epoch": 10.675718849840255, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 13366 + }, + { + "epoch": 10.676517571884984, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 13367 + }, + { + "epoch": 10.677316293929712, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 13368 + }, + { + "epoch": 10.678115015974441, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13369 + }, + { + "epoch": 10.678913738019169, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 13370 + }, + { + "epoch": 10.679712460063898, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 13371 + }, + { + "epoch": 10.680511182108626, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 13372 + }, + { + "epoch": 10.681309904153355, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 13373 + }, + { + "epoch": 10.682108626198083, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 13374 + }, + { + "epoch": 10.68290734824281, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 13375 + }, + { + "epoch": 10.68370607028754, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 13376 + }, + { + "epoch": 10.684504792332268, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 13377 + }, + { + "epoch": 10.685303514376997, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 13378 + }, + { + "epoch": 10.686102236421725, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 13379 + }, + { + "epoch": 10.686900958466454, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 13380 + }, + { + "epoch": 10.687699680511182, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 13381 + }, + { + "epoch": 10.688498402555911, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 13382 + }, + { + "epoch": 10.689297124600639, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 13383 + }, + { + "epoch": 10.690095846645367, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 13384 + }, + { + "epoch": 10.690894568690096, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 13385 + }, + { + "epoch": 10.691693290734824, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 13386 + }, + { + "epoch": 10.692492012779553, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 13387 + }, + { + "epoch": 10.69329073482428, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 13388 + }, + { + "epoch": 10.69408945686901, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 13389 + }, + { + "epoch": 10.694888178913738, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 13390 + }, + { + "epoch": 10.695686900958467, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 13391 + }, + { + "epoch": 10.696485623003195, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 13392 + }, + { + "epoch": 10.697284345047922, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 13393 + }, + { + "epoch": 10.698083067092652, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 13394 + }, + { + "epoch": 10.69888178913738, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 13395 + }, + { + "epoch": 10.699680511182109, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 13396 + }, + { + "epoch": 10.700479233226837, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 13397 + }, + { + "epoch": 10.701277955271566, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 13398 + }, + { + "epoch": 10.702076677316294, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 13399 + }, + { + "epoch": 10.702875399361023, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 13400 + }, + { + "epoch": 10.70367412140575, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 13401 + }, + { + "epoch": 10.704472843450478, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 13402 + }, + { + "epoch": 10.705271565495208, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 13403 + }, + { + "epoch": 10.706070287539935, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 13404 + }, + { + "epoch": 10.706869009584665, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 13405 + }, + { + "epoch": 10.707667731629392, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 13406 + }, + { + "epoch": 10.708466453674122, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 13407 + }, + { + "epoch": 10.70926517571885, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 13408 + }, + { + "epoch": 10.710063897763579, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 13409 + }, + { + "epoch": 10.710862619808307, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 13410 + }, + { + "epoch": 10.711661341853034, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 13411 + }, + { + "epoch": 10.712460063897764, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 13412 + }, + { + "epoch": 10.713258785942491, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13413 + }, + { + "epoch": 10.71405750798722, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 13414 + }, + { + "epoch": 10.714856230031948, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 13415 + }, + { + "epoch": 10.715654952076678, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 13416 + }, + { + "epoch": 10.716453674121405, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 13417 + }, + { + "epoch": 10.717252396166135, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 13418 + }, + { + "epoch": 10.718051118210862, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 13419 + }, + { + "epoch": 10.718849840255592, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 13420 + }, + { + "epoch": 10.71964856230032, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 13421 + }, + { + "epoch": 10.720447284345047, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 13422 + }, + { + "epoch": 10.721246006389777, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 13423 + }, + { + "epoch": 10.722044728434504, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 13424 + }, + { + "epoch": 10.722843450479234, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 13425 + }, + { + "epoch": 10.723642172523961, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 13426 + }, + { + "epoch": 10.72444089456869, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 13427 + }, + { + "epoch": 10.725239616613418, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 13428 + }, + { + "epoch": 10.726038338658148, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 13429 + }, + { + "epoch": 10.726837060702875, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 13430 + }, + { + "epoch": 10.727635782747605, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 13431 + }, + { + "epoch": 10.728434504792332, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 13432 + }, + { + "epoch": 10.72923322683706, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 13433 + }, + { + "epoch": 10.73003194888179, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 13434 + }, + { + "epoch": 10.730830670926517, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 13435 + }, + { + "epoch": 10.731629392971247, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 13436 + }, + { + "epoch": 10.732428115015974, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 13437 + }, + { + "epoch": 10.733226837060704, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 13438 + }, + { + "epoch": 10.734025559105431, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 13439 + }, + { + "epoch": 10.73482428115016, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 13440 + }, + { + "epoch": 10.735623003194888, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 13441 + }, + { + "epoch": 10.736421725239616, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 13442 + }, + { + "epoch": 10.737220447284345, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13443 + }, + { + "epoch": 10.738019169329073, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 13444 + }, + { + "epoch": 10.738817891373802, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 13445 + }, + { + "epoch": 10.73961661341853, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 13446 + }, + { + "epoch": 10.74041533546326, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 13447 + }, + { + "epoch": 10.741214057507987, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 13448 + }, + { + "epoch": 10.742012779552716, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13449 + }, + { + "epoch": 10.742811501597444, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 13450 + }, + { + "epoch": 10.743610223642172, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 13451 + }, + { + "epoch": 10.744408945686901, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 13452 + }, + { + "epoch": 10.745207667731629, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 13453 + }, + { + "epoch": 10.746006389776358, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 13454 + }, + { + "epoch": 10.746805111821086, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13455 + }, + { + "epoch": 10.747603833865815, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 13456 + }, + { + "epoch": 10.748402555910543, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 13457 + }, + { + "epoch": 10.749201277955272, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13458 + }, + { + "epoch": 10.75, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 13459 + }, + { + "epoch": 10.750798722044728, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 13460 + }, + { + "epoch": 10.751597444089457, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 13461 + }, + { + "epoch": 10.752396166134185, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 13462 + }, + { + "epoch": 10.753194888178914, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13463 + }, + { + "epoch": 10.753993610223642, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 13464 + }, + { + "epoch": 10.754792332268371, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 13465 + }, + { + "epoch": 10.755591054313099, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 13466 + }, + { + "epoch": 10.756389776357828, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13467 + }, + { + "epoch": 10.757188498402556, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 13468 + }, + { + "epoch": 10.757987220447284, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13469 + }, + { + "epoch": 10.758785942492013, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 13470 + }, + { + "epoch": 10.75958466453674, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 13471 + }, + { + "epoch": 10.76038338658147, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 13472 + }, + { + "epoch": 10.761182108626198, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 13473 + }, + { + "epoch": 10.761980830670927, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 13474 + }, + { + "epoch": 10.762779552715655, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 13475 + }, + { + "epoch": 10.763578274760384, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 13476 + }, + { + "epoch": 10.764376996805112, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 13477 + }, + { + "epoch": 10.76517571884984, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 13478 + }, + { + "epoch": 10.765974440894569, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 13479 + }, + { + "epoch": 10.766773162939296, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 13480 + }, + { + "epoch": 10.767571884984026, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 13481 + }, + { + "epoch": 10.768370607028753, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 13482 + }, + { + "epoch": 10.769169329073483, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 13483 + }, + { + "epoch": 10.76996805111821, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 13484 + }, + { + "epoch": 10.77076677316294, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 13485 + }, + { + "epoch": 10.771565495207668, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 13486 + }, + { + "epoch": 10.772364217252395, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 13487 + }, + { + "epoch": 10.773162939297125, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 13488 + }, + { + "epoch": 10.773961661341852, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 13489 + }, + { + "epoch": 10.774760383386582, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 13490 + }, + { + "epoch": 10.77555910543131, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 13491 + }, + { + "epoch": 10.776357827476039, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 13492 + }, + { + "epoch": 10.777156549520766, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 13493 + }, + { + "epoch": 10.777955271565496, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 13494 + }, + { + "epoch": 10.778753993610223, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 13495 + }, + { + "epoch": 10.779552715654953, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 13496 + }, + { + "epoch": 10.78035143769968, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 13497 + }, + { + "epoch": 10.781150159744408, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 13498 + }, + { + "epoch": 10.781948881789138, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 13499 + }, + { + "epoch": 10.782747603833865, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 13500 + }, + { + "epoch": 10.783546325878595, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 13501 + }, + { + "epoch": 10.784345047923322, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 13502 + }, + { + "epoch": 10.785143769968052, + "grad_norm": 0.0966796875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 13503 + }, + { + "epoch": 10.78594249201278, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 13504 + }, + { + "epoch": 10.786741214057509, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 13505 + }, + { + "epoch": 10.787539936102236, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 13506 + }, + { + "epoch": 10.788338658146966, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 13507 + }, + { + "epoch": 10.789137380191693, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 13508 + }, + { + "epoch": 10.789936102236421, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 13509 + }, + { + "epoch": 10.79073482428115, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 13510 + }, + { + "epoch": 10.791533546325878, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 13511 + }, + { + "epoch": 10.792332268370608, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13512 + }, + { + "epoch": 10.793130990415335, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 13513 + }, + { + "epoch": 10.793929712460065, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 13514 + }, + { + "epoch": 10.794728434504792, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 13515 + }, + { + "epoch": 10.795527156549522, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 13516 + }, + { + "epoch": 10.79632587859425, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 13517 + }, + { + "epoch": 10.797124600638977, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 13518 + }, + { + "epoch": 10.797923322683706, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 13519 + }, + { + "epoch": 10.798722044728434, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 13520 + }, + { + "epoch": 10.799520766773163, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 13521 + }, + { + "epoch": 10.800319488817891, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 13522 + }, + { + "epoch": 10.80111821086262, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 13523 + }, + { + "epoch": 10.801916932907348, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 13524 + }, + { + "epoch": 10.802715654952078, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 13525 + }, + { + "epoch": 10.803514376996805, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 13526 + }, + { + "epoch": 10.804313099041533, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 13527 + }, + { + "epoch": 10.805111821086262, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 13528 + }, + { + "epoch": 10.80591054313099, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 13529 + }, + { + "epoch": 10.80670926517572, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 13530 + }, + { + "epoch": 10.807507987220447, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 13531 + }, + { + "epoch": 10.808306709265176, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 13532 + }, + { + "epoch": 10.809105431309904, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 13533 + }, + { + "epoch": 10.809904153354633, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 13534 + }, + { + "epoch": 10.810702875399361, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 13535 + }, + { + "epoch": 10.811501597444089, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 13536 + }, + { + "epoch": 10.812300319488818, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 13537 + }, + { + "epoch": 10.813099041533546, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 13538 + }, + { + "epoch": 10.813897763578275, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 13539 + }, + { + "epoch": 10.814696485623003, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 13540 + }, + { + "epoch": 10.815495207667732, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 13541 + }, + { + "epoch": 10.81629392971246, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 13542 + }, + { + "epoch": 10.81709265175719, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 13543 + }, + { + "epoch": 10.817891373801917, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 13544 + }, + { + "epoch": 10.818690095846645, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 13545 + }, + { + "epoch": 10.819488817891374, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 13546 + }, + { + "epoch": 10.820287539936102, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 13547 + }, + { + "epoch": 10.821086261980831, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 13548 + }, + { + "epoch": 10.821884984025559, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 13549 + }, + { + "epoch": 10.822683706070288, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 13550 + }, + { + "epoch": 10.823482428115016, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 13551 + }, + { + "epoch": 10.824281150159745, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 13552 + }, + { + "epoch": 10.825079872204473, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 13553 + }, + { + "epoch": 10.8258785942492, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 13554 + }, + { + "epoch": 10.82667731629393, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 13555 + }, + { + "epoch": 10.827476038338657, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 13556 + }, + { + "epoch": 10.828274760383387, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 13557 + }, + { + "epoch": 10.829073482428115, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 13558 + }, + { + "epoch": 10.829872204472844, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 13559 + }, + { + "epoch": 10.830670926517572, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 13560 + }, + { + "epoch": 10.831469648562301, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 13561 + }, + { + "epoch": 10.832268370607029, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 13562 + }, + { + "epoch": 10.833067092651756, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 13563 + }, + { + "epoch": 10.833865814696486, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 13564 + }, + { + "epoch": 10.834664536741213, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 13565 + }, + { + "epoch": 10.835463258785943, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 13566 + }, + { + "epoch": 10.83626198083067, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 13567 + }, + { + "epoch": 10.8370607028754, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 13568 + }, + { + "epoch": 10.837859424920127, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 13569 + }, + { + "epoch": 10.838658146964857, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 13570 + }, + { + "epoch": 10.839456869009584, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 13571 + }, + { + "epoch": 10.840255591054314, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 13572 + }, + { + "epoch": 10.841054313099042, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 13573 + }, + { + "epoch": 10.84185303514377, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 13574 + }, + { + "epoch": 10.842651757188499, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 13575 + }, + { + "epoch": 10.843450479233226, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 13576 + }, + { + "epoch": 10.844249201277956, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 13577 + }, + { + "epoch": 10.845047923322683, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 13578 + }, + { + "epoch": 10.845846645367413, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 13579 + }, + { + "epoch": 10.84664536741214, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 13580 + }, + { + "epoch": 10.84744408945687, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 13581 + }, + { + "epoch": 10.848242811501597, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 13582 + }, + { + "epoch": 10.849041533546325, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 13583 + }, + { + "epoch": 10.849840255591054, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 13584 + }, + { + "epoch": 10.850638977635782, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 13585 + }, + { + "epoch": 10.851437699680512, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 13586 + }, + { + "epoch": 10.85223642172524, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 13587 + }, + { + "epoch": 10.853035143769969, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 13588 + }, + { + "epoch": 10.853833865814696, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 13589 + }, + { + "epoch": 10.854632587859426, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 13590 + }, + { + "epoch": 10.855431309904153, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 13591 + }, + { + "epoch": 10.856230031948883, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 13592 + }, + { + "epoch": 10.85702875399361, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 13593 + }, + { + "epoch": 10.857827476038338, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 13594 + }, + { + "epoch": 10.858626198083067, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 13595 + }, + { + "epoch": 10.859424920127795, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 13596 + }, + { + "epoch": 10.860223642172524, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 13597 + }, + { + "epoch": 10.861022364217252, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 13598 + }, + { + "epoch": 10.861821086261982, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 13599 + }, + { + "epoch": 10.86261980830671, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 13600 + }, + { + "epoch": 10.863418530351439, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 13601 + }, + { + "epoch": 10.864217252396166, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 13602 + }, + { + "epoch": 10.865015974440894, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 13603 + }, + { + "epoch": 10.865814696485623, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 13604 + }, + { + "epoch": 10.866613418530351, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 13605 + }, + { + "epoch": 10.86741214057508, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 13606 + }, + { + "epoch": 10.868210862619808, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 13607 + }, + { + "epoch": 10.869009584664537, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 13608 + }, + { + "epoch": 10.869808306709265, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 13609 + }, + { + "epoch": 10.870607028753994, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 13610 + }, + { + "epoch": 10.871405750798722, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 13611 + }, + { + "epoch": 10.87220447284345, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 13612 + }, + { + "epoch": 10.87300319488818, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 13613 + }, + { + "epoch": 10.873801916932907, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 13614 + }, + { + "epoch": 10.874600638977636, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13615 + }, + { + "epoch": 10.875399361022364, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 13616 + }, + { + "epoch": 10.876198083067093, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 13617 + }, + { + "epoch": 10.87699680511182, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 13618 + }, + { + "epoch": 10.87779552715655, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 13619 + }, + { + "epoch": 10.878594249201278, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 13620 + }, + { + "epoch": 10.879392971246006, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 13621 + }, + { + "epoch": 10.880191693290735, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 13622 + }, + { + "epoch": 10.880990415335463, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 13623 + }, + { + "epoch": 10.881789137380192, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 13624 + }, + { + "epoch": 10.88258785942492, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 13625 + }, + { + "epoch": 10.883386581469649, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 13626 + }, + { + "epoch": 10.884185303514377, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 13627 + }, + { + "epoch": 10.884984025559106, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13628 + }, + { + "epoch": 10.885782747603834, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 13629 + }, + { + "epoch": 10.886581469648561, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 13630 + }, + { + "epoch": 10.88738019169329, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 13631 + }, + { + "epoch": 10.888178913738018, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 13632 + }, + { + "epoch": 10.888977635782748, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 13633 + }, + { + "epoch": 10.889776357827476, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 13634 + }, + { + "epoch": 10.890575079872205, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 13635 + }, + { + "epoch": 10.891373801916933, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 13636 + }, + { + "epoch": 10.892172523961662, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 13637 + }, + { + "epoch": 10.89297124600639, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 13638 + }, + { + "epoch": 10.893769968051117, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13639 + }, + { + "epoch": 10.894568690095847, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 13640 + }, + { + "epoch": 10.895367412140574, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 13641 + }, + { + "epoch": 10.896166134185304, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13642 + }, + { + "epoch": 10.896964856230031, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 13643 + }, + { + "epoch": 10.89776357827476, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 13644 + }, + { + "epoch": 10.898562300319488, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 13645 + }, + { + "epoch": 10.899361022364218, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 13646 + }, + { + "epoch": 10.900159744408946, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 13647 + }, + { + "epoch": 10.900958466453675, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 13648 + }, + { + "epoch": 10.901757188498403, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 13649 + }, + { + "epoch": 10.90255591054313, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 13650 + }, + { + "epoch": 10.90335463258786, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 13651 + }, + { + "epoch": 10.904153354632587, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 13652 + }, + { + "epoch": 10.904952076677317, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 13653 + }, + { + "epoch": 10.905750798722044, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 13654 + }, + { + "epoch": 10.906549520766774, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 13655 + }, + { + "epoch": 10.907348242811501, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 13656 + }, + { + "epoch": 10.90814696485623, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 13657 + }, + { + "epoch": 10.908945686900958, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 13658 + }, + { + "epoch": 10.909744408945686, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 13659 + }, + { + "epoch": 10.910543130990416, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 13660 + }, + { + "epoch": 10.911341853035143, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 13661 + }, + { + "epoch": 10.912140575079873, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 13662 + }, + { + "epoch": 10.9129392971246, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 13663 + }, + { + "epoch": 10.91373801916933, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 13664 + }, + { + "epoch": 10.914536741214057, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 13665 + }, + { + "epoch": 10.915335463258787, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 13666 + }, + { + "epoch": 10.916134185303514, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 13667 + }, + { + "epoch": 10.916932907348244, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 13668 + }, + { + "epoch": 10.917731629392971, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 13669 + }, + { + "epoch": 10.918530351437699, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 13670 + }, + { + "epoch": 10.919329073482428, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 13671 + }, + { + "epoch": 10.920127795527156, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 13672 + }, + { + "epoch": 10.920926517571885, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13673 + }, + { + "epoch": 10.921725239616613, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 13674 + }, + { + "epoch": 10.922523961661343, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 13675 + }, + { + "epoch": 10.92332268370607, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 13676 + }, + { + "epoch": 10.9241214057508, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 13677 + }, + { + "epoch": 10.924920127795527, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 13678 + }, + { + "epoch": 10.925718849840255, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 13679 + }, + { + "epoch": 10.926517571884984, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 13680 + }, + { + "epoch": 10.927316293929712, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 13681 + }, + { + "epoch": 10.928115015974441, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 13682 + }, + { + "epoch": 10.928913738019169, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13683 + }, + { + "epoch": 10.929712460063898, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1108, + "step": 13684 + }, + { + "epoch": 10.930511182108626, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 13685 + }, + { + "epoch": 10.931309904153355, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 13686 + }, + { + "epoch": 10.932108626198083, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 13687 + }, + { + "epoch": 10.93290734824281, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 13688 + }, + { + "epoch": 10.93370607028754, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 13689 + }, + { + "epoch": 10.934504792332268, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 13690 + }, + { + "epoch": 10.935303514376997, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 13691 + }, + { + "epoch": 10.936102236421725, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 13692 + }, + { + "epoch": 10.936900958466454, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 13693 + }, + { + "epoch": 10.937699680511182, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 13694 + }, + { + "epoch": 10.938498402555911, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 13695 + }, + { + "epoch": 10.939297124600639, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 13696 + }, + { + "epoch": 10.940095846645367, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 13697 + }, + { + "epoch": 10.940894568690096, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 13698 + }, + { + "epoch": 10.941693290734824, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 13699 + }, + { + "epoch": 10.942492012779553, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 13700 + }, + { + "epoch": 10.94329073482428, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 13701 + }, + { + "epoch": 10.94408945686901, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 13702 + }, + { + "epoch": 10.944888178913738, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 13703 + }, + { + "epoch": 10.945686900958467, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 13704 + }, + { + "epoch": 10.946485623003195, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 13705 + }, + { + "epoch": 10.947284345047922, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 13706 + }, + { + "epoch": 10.948083067092652, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 13707 + }, + { + "epoch": 10.94888178913738, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 13708 + }, + { + "epoch": 10.949680511182109, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 13709 + }, + { + "epoch": 10.950479233226837, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1153, + "step": 13710 + }, + { + "epoch": 10.951277955271566, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 13711 + }, + { + "epoch": 10.952076677316294, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 13712 + }, + { + "epoch": 10.952875399361023, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 13713 + }, + { + "epoch": 10.95367412140575, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 13714 + }, + { + "epoch": 10.954472843450478, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 13715 + }, + { + "epoch": 10.955271565495208, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 13716 + }, + { + "epoch": 10.956070287539935, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 13717 + }, + { + "epoch": 10.956869009584665, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 13718 + }, + { + "epoch": 10.957667731629392, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 13719 + }, + { + "epoch": 10.958466453674122, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 13720 + }, + { + "epoch": 10.95926517571885, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 13721 + }, + { + "epoch": 10.960063897763579, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 13722 + }, + { + "epoch": 10.960862619808307, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 13723 + }, + { + "epoch": 10.961661341853034, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 13724 + }, + { + "epoch": 10.962460063897764, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 13725 + }, + { + "epoch": 10.963258785942491, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 13726 + }, + { + "epoch": 10.96405750798722, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 13727 + }, + { + "epoch": 10.964856230031948, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 13728 + }, + { + "epoch": 10.965654952076678, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 13729 + }, + { + "epoch": 10.966453674121405, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 13730 + }, + { + "epoch": 10.967252396166135, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 13731 + }, + { + "epoch": 10.968051118210862, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 13732 + }, + { + "epoch": 10.968849840255592, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13733 + }, + { + "epoch": 10.96964856230032, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 13734 + }, + { + "epoch": 10.970447284345047, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 13735 + }, + { + "epoch": 10.971246006389777, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 13736 + }, + { + "epoch": 10.972044728434504, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 13737 + }, + { + "epoch": 10.972843450479234, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 13738 + }, + { + "epoch": 10.973642172523961, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 13739 + }, + { + "epoch": 10.97444089456869, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 13740 + }, + { + "epoch": 10.975239616613418, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 13741 + }, + { + "epoch": 10.976038338658148, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 13742 + }, + { + "epoch": 10.976837060702875, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1137, + "step": 13743 + }, + { + "epoch": 10.977635782747605, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 13744 + }, + { + "epoch": 10.978434504792332, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 13745 + }, + { + "epoch": 10.97923322683706, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 13746 + }, + { + "epoch": 10.98003194888179, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 13747 + }, + { + "epoch": 10.980830670926517, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 13748 + }, + { + "epoch": 10.981629392971247, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 13749 + }, + { + "epoch": 10.982428115015974, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 13750 + }, + { + "epoch": 10.983226837060704, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 13751 + }, + { + "epoch": 10.984025559105431, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 13752 + }, + { + "epoch": 10.98482428115016, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 13753 + }, + { + "epoch": 10.985623003194888, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 13754 + }, + { + "epoch": 10.986421725239616, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13755 + }, + { + "epoch": 10.987220447284345, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 13756 + }, + { + "epoch": 10.988019169329073, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 13757 + }, + { + "epoch": 10.988817891373802, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 13758 + }, + { + "epoch": 10.98961661341853, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 13759 + }, + { + "epoch": 10.99041533546326, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 13760 + }, + { + "epoch": 10.991214057507987, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 13761 + }, + { + "epoch": 10.992012779552716, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 13762 + }, + { + "epoch": 10.992811501597444, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 13763 + }, + { + "epoch": 10.993610223642172, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 13764 + }, + { + "epoch": 10.994408945686901, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 13765 + }, + { + "epoch": 10.995207667731629, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13766 + }, + { + "epoch": 10.996006389776358, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 13767 + }, + { + "epoch": 10.996805111821086, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 13768 + }, + { + "epoch": 10.997603833865815, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 13769 + }, + { + "epoch": 10.998402555910543, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 13770 + }, + { + "epoch": 10.999201277955272, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 13771 + }, + { + "epoch": 11.0, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 13772 + }, + { + "epoch": 11.000798722044728, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 13773 + }, + { + "epoch": 11.001597444089457, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 13774 + }, + { + "epoch": 11.002396166134185, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 13775 + }, + { + "epoch": 11.003194888178914, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13776 + }, + { + "epoch": 11.003993610223642, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 13777 + }, + { + "epoch": 11.004792332268371, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13778 + }, + { + "epoch": 11.005591054313099, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 13779 + }, + { + "epoch": 11.006389776357828, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 13780 + }, + { + "epoch": 11.007188498402556, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 13781 + }, + { + "epoch": 11.007987220447284, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 13782 + }, + { + "epoch": 11.008785942492013, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 13783 + }, + { + "epoch": 11.00958466453674, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 13784 + }, + { + "epoch": 11.01038338658147, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 13785 + }, + { + "epoch": 11.011182108626198, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 13786 + }, + { + "epoch": 11.011980830670927, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 13787 + }, + { + "epoch": 11.012779552715655, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 13788 + }, + { + "epoch": 11.013578274760384, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 13789 + }, + { + "epoch": 11.014376996805112, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 13790 + }, + { + "epoch": 11.01517571884984, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 13791 + }, + { + "epoch": 11.015974440894569, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 13792 + }, + { + "epoch": 11.016773162939296, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 13793 + }, + { + "epoch": 11.017571884984026, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 13794 + }, + { + "epoch": 11.018370607028753, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13795 + }, + { + "epoch": 11.019169329073483, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 13796 + }, + { + "epoch": 11.01996805111821, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13797 + }, + { + "epoch": 11.02076677316294, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 13798 + }, + { + "epoch": 11.021565495207668, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 13799 + }, + { + "epoch": 11.022364217252397, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 13800 + }, + { + "epoch": 11.023162939297125, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 13801 + }, + { + "epoch": 11.023961661341852, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 13802 + }, + { + "epoch": 11.024760383386582, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 13803 + }, + { + "epoch": 11.02555910543131, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 13804 + }, + { + "epoch": 11.026357827476039, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 13805 + }, + { + "epoch": 11.027156549520766, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 13806 + }, + { + "epoch": 11.027955271565496, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 13807 + }, + { + "epoch": 11.028753993610223, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 13808 + }, + { + "epoch": 11.029552715654953, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 13809 + }, + { + "epoch": 11.03035143769968, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 13810 + }, + { + "epoch": 11.031150159744408, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 13811 + }, + { + "epoch": 11.031948881789138, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 13812 + }, + { + "epoch": 11.032747603833865, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 13813 + }, + { + "epoch": 11.033546325878595, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 13814 + }, + { + "epoch": 11.034345047923322, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 13815 + }, + { + "epoch": 11.035143769968052, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 13816 + }, + { + "epoch": 11.03594249201278, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13817 + }, + { + "epoch": 11.036741214057509, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 13818 + }, + { + "epoch": 11.037539936102236, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 13819 + }, + { + "epoch": 11.038338658146964, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 13820 + }, + { + "epoch": 11.039137380191693, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 13821 + }, + { + "epoch": 11.039936102236421, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 13822 + }, + { + "epoch": 11.04073482428115, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 13823 + }, + { + "epoch": 11.041533546325878, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 13824 + }, + { + "epoch": 11.042332268370608, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 13825 + }, + { + "epoch": 11.043130990415335, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 13826 + }, + { + "epoch": 11.043929712460065, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 13827 + }, + { + "epoch": 11.044728434504792, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13828 + }, + { + "epoch": 11.04552715654952, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 13829 + }, + { + "epoch": 11.04632587859425, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 13830 + }, + { + "epoch": 11.047124600638977, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 13831 + }, + { + "epoch": 11.047923322683706, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 13832 + }, + { + "epoch": 11.048722044728434, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 13833 + }, + { + "epoch": 11.049520766773163, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 13834 + }, + { + "epoch": 11.050319488817891, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13835 + }, + { + "epoch": 11.05111821086262, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 13836 + }, + { + "epoch": 11.051916932907348, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 13837 + }, + { + "epoch": 11.052715654952078, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 13838 + }, + { + "epoch": 11.053514376996805, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 13839 + }, + { + "epoch": 11.054313099041533, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 13840 + }, + { + "epoch": 11.055111821086262, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 13841 + }, + { + "epoch": 11.05591054313099, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 13842 + }, + { + "epoch": 11.05670926517572, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 13843 + }, + { + "epoch": 11.057507987220447, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 13844 + }, + { + "epoch": 11.058306709265176, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13845 + }, + { + "epoch": 11.059105431309904, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 13846 + }, + { + "epoch": 11.059904153354633, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 13847 + }, + { + "epoch": 11.060702875399361, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 13848 + }, + { + "epoch": 11.061501597444089, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 13849 + }, + { + "epoch": 11.062300319488818, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 13850 + }, + { + "epoch": 11.063099041533546, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 13851 + }, + { + "epoch": 11.063897763578275, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 13852 + }, + { + "epoch": 11.064696485623003, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 13853 + }, + { + "epoch": 11.065495207667732, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 13854 + }, + { + "epoch": 11.06629392971246, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 13855 + }, + { + "epoch": 11.06709265175719, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 13856 + }, + { + "epoch": 11.067891373801917, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 13857 + }, + { + "epoch": 11.068690095846645, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 13858 + }, + { + "epoch": 11.069488817891374, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 13859 + }, + { + "epoch": 11.070287539936102, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 13860 + }, + { + "epoch": 11.071086261980831, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 13861 + }, + { + "epoch": 11.071884984025559, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 13862 + }, + { + "epoch": 11.072683706070288, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 13863 + }, + { + "epoch": 11.073482428115016, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 13864 + }, + { + "epoch": 11.074281150159745, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 13865 + }, + { + "epoch": 11.075079872204473, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 13866 + }, + { + "epoch": 11.0758785942492, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 13867 + }, + { + "epoch": 11.07667731629393, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 13868 + }, + { + "epoch": 11.077476038338657, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13869 + }, + { + "epoch": 11.078274760383387, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 13870 + }, + { + "epoch": 11.079073482428115, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 13871 + }, + { + "epoch": 11.079872204472844, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 13872 + }, + { + "epoch": 11.080670926517572, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 13873 + }, + { + "epoch": 11.081469648562301, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 13874 + }, + { + "epoch": 11.082268370607029, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 13875 + }, + { + "epoch": 11.083067092651758, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 13876 + }, + { + "epoch": 11.083865814696486, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 13877 + }, + { + "epoch": 11.084664536741213, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 13878 + }, + { + "epoch": 11.085463258785943, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 13879 + }, + { + "epoch": 11.08626198083067, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 13880 + }, + { + "epoch": 11.0870607028754, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 13881 + }, + { + "epoch": 11.087859424920127, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 13882 + }, + { + "epoch": 11.088658146964857, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 13883 + }, + { + "epoch": 11.089456869009584, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 13884 + }, + { + "epoch": 11.090255591054314, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 13885 + }, + { + "epoch": 11.091054313099042, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 13886 + }, + { + "epoch": 11.09185303514377, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 13887 + }, + { + "epoch": 11.092651757188499, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 13888 + }, + { + "epoch": 11.093450479233226, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 13889 + }, + { + "epoch": 11.094249201277956, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 13890 + }, + { + "epoch": 11.095047923322683, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 13891 + }, + { + "epoch": 11.095846645367413, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 13892 + }, + { + "epoch": 11.09664536741214, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 13893 + }, + { + "epoch": 11.09744408945687, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 13894 + }, + { + "epoch": 11.098242811501597, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 13895 + }, + { + "epoch": 11.099041533546325, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 13896 + }, + { + "epoch": 11.099840255591054, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 13897 + }, + { + "epoch": 11.100638977635782, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 13898 + }, + { + "epoch": 11.101437699680512, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 13899 + }, + { + "epoch": 11.10223642172524, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 13900 + }, + { + "epoch": 11.103035143769969, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 13901 + }, + { + "epoch": 11.103833865814696, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 13902 + }, + { + "epoch": 11.104632587859426, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 13903 + }, + { + "epoch": 11.105431309904153, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 13904 + }, + { + "epoch": 11.106230031948881, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 13905 + }, + { + "epoch": 11.10702875399361, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 13906 + }, + { + "epoch": 11.107827476038338, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 13907 + }, + { + "epoch": 11.108626198083067, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 13908 + }, + { + "epoch": 11.109424920127795, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 13909 + }, + { + "epoch": 11.110223642172524, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 13910 + }, + { + "epoch": 11.111022364217252, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 13911 + }, + { + "epoch": 11.111821086261982, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 13912 + }, + { + "epoch": 11.11261980830671, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 13913 + }, + { + "epoch": 11.113418530351439, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 13914 + }, + { + "epoch": 11.114217252396166, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13915 + }, + { + "epoch": 11.115015974440894, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 13916 + }, + { + "epoch": 11.115814696485623, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 13917 + }, + { + "epoch": 11.116613418530351, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 13918 + }, + { + "epoch": 11.11741214057508, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 13919 + }, + { + "epoch": 11.118210862619808, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 13920 + }, + { + "epoch": 11.119009584664537, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 13921 + }, + { + "epoch": 11.119808306709265, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 13922 + }, + { + "epoch": 11.120607028753994, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 13923 + }, + { + "epoch": 11.121405750798722, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 13924 + }, + { + "epoch": 11.12220447284345, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 13925 + }, + { + "epoch": 11.12300319488818, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 13926 + }, + { + "epoch": 11.123801916932907, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13927 + }, + { + "epoch": 11.124600638977636, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 13928 + }, + { + "epoch": 11.125399361022364, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 13929 + }, + { + "epoch": 11.126198083067093, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 13930 + }, + { + "epoch": 11.12699680511182, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 13931 + }, + { + "epoch": 11.12779552715655, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 13932 + }, + { + "epoch": 11.128594249201278, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 13933 + }, + { + "epoch": 11.129392971246006, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 13934 + }, + { + "epoch": 11.130191693290735, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 13935 + }, + { + "epoch": 11.130990415335463, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 13936 + }, + { + "epoch": 11.131789137380192, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 13937 + }, + { + "epoch": 11.13258785942492, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 13938 + }, + { + "epoch": 11.133386581469649, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 13939 + }, + { + "epoch": 11.134185303514377, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 13940 + }, + { + "epoch": 11.134984025559106, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 13941 + }, + { + "epoch": 11.135782747603834, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 13942 + }, + { + "epoch": 11.136581469648561, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 13943 + }, + { + "epoch": 11.13738019169329, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 13944 + }, + { + "epoch": 11.138178913738018, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 13945 + }, + { + "epoch": 11.138977635782748, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 13946 + }, + { + "epoch": 11.139776357827476, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13947 + }, + { + "epoch": 11.140575079872205, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 13948 + }, + { + "epoch": 11.141373801916933, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 13949 + }, + { + "epoch": 11.142172523961662, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 13950 + }, + { + "epoch": 11.14297124600639, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 13951 + }, + { + "epoch": 11.143769968051119, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 13952 + }, + { + "epoch": 11.144568690095847, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 13953 + }, + { + "epoch": 11.145367412140574, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 13954 + }, + { + "epoch": 11.146166134185304, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 13955 + }, + { + "epoch": 11.146964856230031, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 13956 + }, + { + "epoch": 11.14776357827476, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 13957 + }, + { + "epoch": 11.148562300319488, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 13958 + }, + { + "epoch": 11.149361022364218, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 13959 + }, + { + "epoch": 11.150159744408946, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 13960 + }, + { + "epoch": 11.150958466453675, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 13961 + }, + { + "epoch": 11.151757188498403, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 13962 + }, + { + "epoch": 11.15255591054313, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 13963 + }, + { + "epoch": 11.15335463258786, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 13964 + }, + { + "epoch": 11.154153354632587, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 13965 + }, + { + "epoch": 11.154952076677317, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 13966 + }, + { + "epoch": 11.155750798722044, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 13967 + }, + { + "epoch": 11.156549520766774, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 13968 + }, + { + "epoch": 11.157348242811501, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 13969 + }, + { + "epoch": 11.15814696485623, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13970 + }, + { + "epoch": 11.158945686900958, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 13971 + }, + { + "epoch": 11.159744408945686, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 13972 + }, + { + "epoch": 11.160543130990416, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 13973 + }, + { + "epoch": 11.161341853035143, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 13974 + }, + { + "epoch": 11.162140575079873, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 13975 + }, + { + "epoch": 11.1629392971246, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 13976 + }, + { + "epoch": 11.16373801916933, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 13977 + }, + { + "epoch": 11.164536741214057, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 13978 + }, + { + "epoch": 11.165335463258787, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 13979 + }, + { + "epoch": 11.166134185303514, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 13980 + }, + { + "epoch": 11.166932907348242, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 13981 + }, + { + "epoch": 11.167731629392971, + "grad_norm": 0.361328125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 13982 + }, + { + "epoch": 11.168530351437699, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 13983 + }, + { + "epoch": 11.169329073482428, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 13984 + }, + { + "epoch": 11.170127795527156, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 13985 + }, + { + "epoch": 11.170926517571885, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 13986 + }, + { + "epoch": 11.171725239616613, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 13987 + }, + { + "epoch": 11.172523961661343, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 13988 + }, + { + "epoch": 11.17332268370607, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 13989 + }, + { + "epoch": 11.1741214057508, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 13990 + }, + { + "epoch": 11.174920127795527, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 13991 + }, + { + "epoch": 11.175718849840255, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 13992 + }, + { + "epoch": 11.176517571884984, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 13993 + }, + { + "epoch": 11.177316293929712, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 13994 + }, + { + "epoch": 11.178115015974441, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 13995 + }, + { + "epoch": 11.178913738019169, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 13996 + }, + { + "epoch": 11.179712460063898, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 13997 + }, + { + "epoch": 11.180511182108626, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 13998 + }, + { + "epoch": 11.181309904153355, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 13999 + }, + { + "epoch": 11.182108626198083, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 14000 + }, + { + "epoch": 11.18290734824281, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 14001 + }, + { + "epoch": 11.18370607028754, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 14002 + }, + { + "epoch": 11.184504792332268, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 14003 + }, + { + "epoch": 11.185303514376997, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14004 + }, + { + "epoch": 11.186102236421725, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 14005 + }, + { + "epoch": 11.186900958466454, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14006 + }, + { + "epoch": 11.187699680511182, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1129, + "step": 14007 + }, + { + "epoch": 11.188498402555911, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 14008 + }, + { + "epoch": 11.189297124600639, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 14009 + }, + { + "epoch": 11.190095846645367, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 14010 + }, + { + "epoch": 11.190894568690096, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 14011 + }, + { + "epoch": 11.191693290734824, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 14012 + }, + { + "epoch": 11.192492012779553, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 14013 + }, + { + "epoch": 11.19329073482428, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 14014 + }, + { + "epoch": 11.19408945686901, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 14015 + }, + { + "epoch": 11.194888178913738, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 14016 + }, + { + "epoch": 11.195686900958467, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 14017 + }, + { + "epoch": 11.196485623003195, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 14018 + }, + { + "epoch": 11.197284345047922, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 14019 + }, + { + "epoch": 11.198083067092652, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 14020 + }, + { + "epoch": 11.19888178913738, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14021 + }, + { + "epoch": 11.199680511182109, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 14022 + }, + { + "epoch": 11.200479233226837, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 14023 + }, + { + "epoch": 11.201277955271566, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 14024 + }, + { + "epoch": 11.202076677316294, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 14025 + }, + { + "epoch": 11.202875399361023, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 14026 + }, + { + "epoch": 11.20367412140575, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 14027 + }, + { + "epoch": 11.204472843450478, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14028 + }, + { + "epoch": 11.205271565495208, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 14029 + }, + { + "epoch": 11.206070287539935, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 14030 + }, + { + "epoch": 11.206869009584665, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 14031 + }, + { + "epoch": 11.207667731629392, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 14032 + }, + { + "epoch": 11.208466453674122, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 14033 + }, + { + "epoch": 11.20926517571885, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14034 + }, + { + "epoch": 11.210063897763579, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 14035 + }, + { + "epoch": 11.210862619808307, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 14036 + }, + { + "epoch": 11.211661341853036, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 14037 + }, + { + "epoch": 11.212460063897764, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 14038 + }, + { + "epoch": 11.213258785942491, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14039 + }, + { + "epoch": 11.21405750798722, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 14040 + }, + { + "epoch": 11.214856230031948, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 14041 + }, + { + "epoch": 11.215654952076678, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 14042 + }, + { + "epoch": 11.216453674121405, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 14043 + }, + { + "epoch": 11.217252396166135, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 14044 + }, + { + "epoch": 11.218051118210862, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 14045 + }, + { + "epoch": 11.218849840255592, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 14046 + }, + { + "epoch": 11.21964856230032, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 14047 + }, + { + "epoch": 11.220447284345047, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 14048 + }, + { + "epoch": 11.221246006389777, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 14049 + }, + { + "epoch": 11.222044728434504, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 14050 + }, + { + "epoch": 11.222843450479234, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 14051 + }, + { + "epoch": 11.223642172523961, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 14052 + }, + { + "epoch": 11.22444089456869, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 14053 + }, + { + "epoch": 11.225239616613418, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 14054 + }, + { + "epoch": 11.226038338658148, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 14055 + }, + { + "epoch": 11.226837060702875, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 14056 + }, + { + "epoch": 11.227635782747603, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 14057 + }, + { + "epoch": 11.228434504792332, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 14058 + }, + { + "epoch": 11.22923322683706, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 14059 + }, + { + "epoch": 11.23003194888179, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 14060 + }, + { + "epoch": 11.230830670926517, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 14061 + }, + { + "epoch": 11.231629392971247, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 14062 + }, + { + "epoch": 11.232428115015974, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 14063 + }, + { + "epoch": 11.233226837060704, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14064 + }, + { + "epoch": 11.234025559105431, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 14065 + }, + { + "epoch": 11.23482428115016, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 14066 + }, + { + "epoch": 11.235623003194888, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 14067 + }, + { + "epoch": 11.236421725239616, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 14068 + }, + { + "epoch": 11.237220447284345, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 14069 + }, + { + "epoch": 11.238019169329073, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 14070 + }, + { + "epoch": 11.238817891373802, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1186, + "step": 14071 + }, + { + "epoch": 11.23961661341853, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 14072 + }, + { + "epoch": 11.24041533546326, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 14073 + }, + { + "epoch": 11.241214057507987, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 14074 + }, + { + "epoch": 11.242012779552716, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 14075 + }, + { + "epoch": 11.242811501597444, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 14076 + }, + { + "epoch": 11.243610223642172, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 14077 + }, + { + "epoch": 11.244408945686901, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 14078 + }, + { + "epoch": 11.245207667731629, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 14079 + }, + { + "epoch": 11.246006389776358, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 14080 + }, + { + "epoch": 11.246805111821086, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 14081 + }, + { + "epoch": 11.247603833865815, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 14082 + }, + { + "epoch": 11.248402555910543, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1135, + "step": 14083 + }, + { + "epoch": 11.249201277955272, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 14084 + }, + { + "epoch": 11.25, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 14085 + }, + { + "epoch": 11.250798722044728, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 14086 + }, + { + "epoch": 11.251597444089457, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 14087 + }, + { + "epoch": 11.252396166134185, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 14088 + }, + { + "epoch": 11.253194888178914, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 14089 + }, + { + "epoch": 11.253993610223642, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 14090 + }, + { + "epoch": 11.254792332268371, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 14091 + }, + { + "epoch": 11.255591054313099, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 14092 + }, + { + "epoch": 11.256389776357828, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 14093 + }, + { + "epoch": 11.257188498402556, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 14094 + }, + { + "epoch": 11.257987220447284, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 14095 + }, + { + "epoch": 11.258785942492013, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 14096 + }, + { + "epoch": 11.25958466453674, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 14097 + }, + { + "epoch": 11.26038338658147, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 14098 + }, + { + "epoch": 11.261182108626198, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 14099 + }, + { + "epoch": 11.261980830670927, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 14100 + }, + { + "epoch": 11.262779552715655, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 14101 + }, + { + "epoch": 11.263578274760384, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 14102 + }, + { + "epoch": 11.264376996805112, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 14103 + }, + { + "epoch": 11.26517571884984, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 14104 + }, + { + "epoch": 11.265974440894569, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 14105 + }, + { + "epoch": 11.266773162939296, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 14106 + }, + { + "epoch": 11.267571884984026, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 14107 + }, + { + "epoch": 11.268370607028753, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 14108 + }, + { + "epoch": 11.269169329073483, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 14109 + }, + { + "epoch": 11.26996805111821, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 14110 + }, + { + "epoch": 11.27076677316294, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 14111 + }, + { + "epoch": 11.271565495207668, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 14112 + }, + { + "epoch": 11.272364217252397, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 14113 + }, + { + "epoch": 11.273162939297125, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 14114 + }, + { + "epoch": 11.273961661341852, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 14115 + }, + { + "epoch": 11.274760383386582, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1132, + "step": 14116 + }, + { + "epoch": 11.27555910543131, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 14117 + }, + { + "epoch": 11.276357827476039, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 14118 + }, + { + "epoch": 11.277156549520766, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 14119 + }, + { + "epoch": 11.277955271565496, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 14120 + }, + { + "epoch": 11.278753993610223, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 14121 + }, + { + "epoch": 11.279552715654953, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 14122 + }, + { + "epoch": 11.28035143769968, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 14123 + }, + { + "epoch": 11.281150159744408, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 14124 + }, + { + "epoch": 11.281948881789138, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 14125 + }, + { + "epoch": 11.282747603833865, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 14126 + }, + { + "epoch": 11.283546325878595, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 14127 + }, + { + "epoch": 11.284345047923322, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 14128 + }, + { + "epoch": 11.285143769968052, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 14129 + }, + { + "epoch": 11.28594249201278, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14130 + }, + { + "epoch": 11.286741214057509, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 14131 + }, + { + "epoch": 11.287539936102236, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 14132 + }, + { + "epoch": 11.288338658146964, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 14133 + }, + { + "epoch": 11.289137380191693, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 14134 + }, + { + "epoch": 11.289936102236421, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 14135 + }, + { + "epoch": 11.29073482428115, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 14136 + }, + { + "epoch": 11.291533546325878, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 14137 + }, + { + "epoch": 11.292332268370608, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 14138 + }, + { + "epoch": 11.293130990415335, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 14139 + }, + { + "epoch": 11.293929712460065, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 14140 + }, + { + "epoch": 11.294728434504792, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 14141 + }, + { + "epoch": 11.295527156549522, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 14142 + }, + { + "epoch": 11.29632587859425, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 14143 + }, + { + "epoch": 11.297124600638977, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 14144 + }, + { + "epoch": 11.297923322683706, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 14145 + }, + { + "epoch": 11.298722044728434, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 14146 + }, + { + "epoch": 11.299520766773163, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 14147 + }, + { + "epoch": 11.300319488817891, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 14148 + }, + { + "epoch": 11.30111821086262, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 14149 + }, + { + "epoch": 11.301916932907348, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 14150 + }, + { + "epoch": 11.302715654952078, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 14151 + }, + { + "epoch": 11.303514376996805, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 14152 + }, + { + "epoch": 11.304313099041533, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 14153 + }, + { + "epoch": 11.305111821086262, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14154 + }, + { + "epoch": 11.30591054313099, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 14155 + }, + { + "epoch": 11.30670926517572, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 14156 + }, + { + "epoch": 11.307507987220447, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 14157 + }, + { + "epoch": 11.308306709265176, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 14158 + }, + { + "epoch": 11.309105431309904, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 14159 + }, + { + "epoch": 11.309904153354633, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 14160 + }, + { + "epoch": 11.310702875399361, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 14161 + }, + { + "epoch": 11.311501597444089, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 14162 + }, + { + "epoch": 11.312300319488818, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 14163 + }, + { + "epoch": 11.313099041533546, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 14164 + }, + { + "epoch": 11.313897763578275, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14165 + }, + { + "epoch": 11.314696485623003, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 14166 + }, + { + "epoch": 11.315495207667732, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 14167 + }, + { + "epoch": 11.31629392971246, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 14168 + }, + { + "epoch": 11.31709265175719, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 14169 + }, + { + "epoch": 11.317891373801917, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 14170 + }, + { + "epoch": 11.318690095846645, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 14171 + }, + { + "epoch": 11.319488817891374, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 14172 + }, + { + "epoch": 11.320287539936102, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 14173 + }, + { + "epoch": 11.321086261980831, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 14174 + }, + { + "epoch": 11.321884984025559, + "grad_norm": 0.34375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 14175 + }, + { + "epoch": 11.322683706070288, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 14176 + }, + { + "epoch": 11.323482428115016, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 14177 + }, + { + "epoch": 11.324281150159745, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 14178 + }, + { + "epoch": 11.325079872204473, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 14179 + }, + { + "epoch": 11.3258785942492, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 14180 + }, + { + "epoch": 11.32667731629393, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 14181 + }, + { + "epoch": 11.327476038338657, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 14182 + }, + { + "epoch": 11.328274760383387, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 14183 + }, + { + "epoch": 11.329073482428115, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 14184 + }, + { + "epoch": 11.329872204472844, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 14185 + }, + { + "epoch": 11.330670926517572, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14186 + }, + { + "epoch": 11.331469648562301, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 14187 + }, + { + "epoch": 11.332268370607029, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 14188 + }, + { + "epoch": 11.333067092651758, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 14189 + }, + { + "epoch": 11.333865814696486, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 14190 + }, + { + "epoch": 11.334664536741213, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 14191 + }, + { + "epoch": 11.335463258785943, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 14192 + }, + { + "epoch": 11.33626198083067, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 14193 + }, + { + "epoch": 11.3370607028754, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 14194 + }, + { + "epoch": 11.337859424920127, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 14195 + }, + { + "epoch": 11.338658146964857, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 14196 + }, + { + "epoch": 11.339456869009584, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 14197 + }, + { + "epoch": 11.340255591054314, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 14198 + }, + { + "epoch": 11.341054313099042, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14199 + }, + { + "epoch": 11.34185303514377, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 14200 + }, + { + "epoch": 11.342651757188499, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 14201 + }, + { + "epoch": 11.343450479233226, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 14202 + }, + { + "epoch": 11.344249201277956, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14203 + }, + { + "epoch": 11.345047923322683, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 14204 + }, + { + "epoch": 11.345846645367413, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 14205 + }, + { + "epoch": 11.34664536741214, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 14206 + }, + { + "epoch": 11.34744408945687, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 14207 + }, + { + "epoch": 11.348242811501597, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 14208 + }, + { + "epoch": 11.349041533546325, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 14209 + }, + { + "epoch": 11.349840255591054, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 14210 + }, + { + "epoch": 11.350638977635782, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 14211 + }, + { + "epoch": 11.351437699680512, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 14212 + }, + { + "epoch": 11.35223642172524, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 14213 + }, + { + "epoch": 11.353035143769969, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 14214 + }, + { + "epoch": 11.353833865814696, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 14215 + }, + { + "epoch": 11.354632587859426, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 14216 + }, + { + "epoch": 11.355431309904153, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 14217 + }, + { + "epoch": 11.356230031948883, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 14218 + }, + { + "epoch": 11.35702875399361, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 14219 + }, + { + "epoch": 11.357827476038338, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 14220 + }, + { + "epoch": 11.358626198083067, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 14221 + }, + { + "epoch": 11.359424920127795, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 14222 + }, + { + "epoch": 11.360223642172524, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 14223 + }, + { + "epoch": 11.361022364217252, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 14224 + }, + { + "epoch": 11.361821086261982, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 14225 + }, + { + "epoch": 11.36261980830671, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 14226 + }, + { + "epoch": 11.363418530351439, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 14227 + }, + { + "epoch": 11.364217252396166, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 14228 + }, + { + "epoch": 11.365015974440894, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 14229 + }, + { + "epoch": 11.365814696485623, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 14230 + }, + { + "epoch": 11.366613418530351, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 14231 + }, + { + "epoch": 11.36741214057508, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 14232 + }, + { + "epoch": 11.368210862619808, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 14233 + }, + { + "epoch": 11.369009584664537, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 14234 + }, + { + "epoch": 11.369808306709265, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 14235 + }, + { + "epoch": 11.370607028753994, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 14236 + }, + { + "epoch": 11.371405750798722, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 14237 + }, + { + "epoch": 11.37220447284345, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 14238 + }, + { + "epoch": 11.37300319488818, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 14239 + }, + { + "epoch": 11.373801916932907, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 14240 + }, + { + "epoch": 11.374600638977636, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 14241 + }, + { + "epoch": 11.375399361022364, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 14242 + }, + { + "epoch": 11.376198083067093, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 14243 + }, + { + "epoch": 11.37699680511182, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 14244 + }, + { + "epoch": 11.37779552715655, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 14245 + }, + { + "epoch": 11.378594249201278, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 14246 + }, + { + "epoch": 11.379392971246006, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 14247 + }, + { + "epoch": 11.380191693290735, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 14248 + }, + { + "epoch": 11.380990415335463, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 14249 + }, + { + "epoch": 11.381789137380192, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 14250 + }, + { + "epoch": 11.38258785942492, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 14251 + }, + { + "epoch": 11.383386581469649, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 14252 + }, + { + "epoch": 11.384185303514377, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 14253 + }, + { + "epoch": 11.384984025559106, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 14254 + }, + { + "epoch": 11.385782747603834, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 14255 + }, + { + "epoch": 11.386581469648561, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 14256 + }, + { + "epoch": 11.38738019169329, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 14257 + }, + { + "epoch": 11.388178913738018, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 14258 + }, + { + "epoch": 11.388977635782748, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 14259 + }, + { + "epoch": 11.389776357827476, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 14260 + }, + { + "epoch": 11.390575079872205, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 14261 + }, + { + "epoch": 11.391373801916933, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 14262 + }, + { + "epoch": 11.392172523961662, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 14263 + }, + { + "epoch": 11.39297124600639, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 14264 + }, + { + "epoch": 11.393769968051117, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 14265 + }, + { + "epoch": 11.394568690095847, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 14266 + }, + { + "epoch": 11.395367412140574, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 14267 + }, + { + "epoch": 11.396166134185304, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14268 + }, + { + "epoch": 11.396964856230031, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 14269 + }, + { + "epoch": 11.39776357827476, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 14270 + }, + { + "epoch": 11.398562300319488, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14271 + }, + { + "epoch": 11.399361022364218, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 14272 + }, + { + "epoch": 11.400159744408946, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 14273 + }, + { + "epoch": 11.400958466453675, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 14274 + }, + { + "epoch": 11.401757188498403, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 14275 + }, + { + "epoch": 11.40255591054313, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14276 + }, + { + "epoch": 11.40335463258786, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 14277 + }, + { + "epoch": 11.404153354632587, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 14278 + }, + { + "epoch": 11.404952076677317, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 14279 + }, + { + "epoch": 11.405750798722044, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 14280 + }, + { + "epoch": 11.406549520766774, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14281 + }, + { + "epoch": 11.407348242811501, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 14282 + }, + { + "epoch": 11.40814696485623, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 14283 + }, + { + "epoch": 11.408945686900958, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 14284 + }, + { + "epoch": 11.409744408945686, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 14285 + }, + { + "epoch": 11.410543130990416, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 14286 + }, + { + "epoch": 11.411341853035143, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 14287 + }, + { + "epoch": 11.412140575079873, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 14288 + }, + { + "epoch": 11.4129392971246, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 14289 + }, + { + "epoch": 11.41373801916933, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 14290 + }, + { + "epoch": 11.414536741214057, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 14291 + }, + { + "epoch": 11.415335463258787, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 14292 + }, + { + "epoch": 11.416134185303514, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 14293 + }, + { + "epoch": 11.416932907348242, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 14294 + }, + { + "epoch": 11.417731629392971, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 14295 + }, + { + "epoch": 11.418530351437699, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 14296 + }, + { + "epoch": 11.419329073482428, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 14297 + }, + { + "epoch": 11.420127795527156, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 14298 + }, + { + "epoch": 11.420926517571885, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14299 + }, + { + "epoch": 11.421725239616613, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 14300 + }, + { + "epoch": 11.422523961661343, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 14301 + }, + { + "epoch": 11.42332268370607, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 14302 + }, + { + "epoch": 11.4241214057508, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 14303 + }, + { + "epoch": 11.424920127795527, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 14304 + }, + { + "epoch": 11.425718849840255, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 14305 + }, + { + "epoch": 11.426517571884984, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 14306 + }, + { + "epoch": 11.427316293929712, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 14307 + }, + { + "epoch": 11.428115015974441, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 14308 + }, + { + "epoch": 11.428913738019169, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 14309 + }, + { + "epoch": 11.429712460063898, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 14310 + }, + { + "epoch": 11.430511182108626, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 14311 + }, + { + "epoch": 11.431309904153355, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 14312 + }, + { + "epoch": 11.432108626198083, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 14313 + }, + { + "epoch": 11.43290734824281, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 14314 + }, + { + "epoch": 11.43370607028754, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 14315 + }, + { + "epoch": 11.434504792332268, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 14316 + }, + { + "epoch": 11.435303514376997, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 14317 + }, + { + "epoch": 11.436102236421725, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 14318 + }, + { + "epoch": 11.436900958466454, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 14319 + }, + { + "epoch": 11.437699680511182, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 14320 + }, + { + "epoch": 11.438498402555911, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 14321 + }, + { + "epoch": 11.439297124600639, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 14322 + }, + { + "epoch": 11.440095846645367, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 14323 + }, + { + "epoch": 11.440894568690096, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 14324 + }, + { + "epoch": 11.441693290734824, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 14325 + }, + { + "epoch": 11.442492012779553, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 14326 + }, + { + "epoch": 11.44329073482428, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 14327 + }, + { + "epoch": 11.44408945686901, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 14328 + }, + { + "epoch": 11.444888178913738, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 14329 + }, + { + "epoch": 11.445686900958467, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 14330 + }, + { + "epoch": 11.446485623003195, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 14331 + }, + { + "epoch": 11.447284345047922, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 14332 + }, + { + "epoch": 11.448083067092652, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 14333 + }, + { + "epoch": 11.44888178913738, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 14334 + }, + { + "epoch": 11.449680511182109, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 14335 + }, + { + "epoch": 11.450479233226837, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 14336 + }, + { + "epoch": 11.451277955271566, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 14337 + }, + { + "epoch": 11.452076677316294, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14338 + }, + { + "epoch": 11.452875399361023, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 14339 + }, + { + "epoch": 11.45367412140575, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 14340 + }, + { + "epoch": 11.454472843450478, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 14341 + }, + { + "epoch": 11.455271565495208, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 14342 + }, + { + "epoch": 11.456070287539935, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 14343 + }, + { + "epoch": 11.456869009584665, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 14344 + }, + { + "epoch": 11.457667731629392, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 14345 + }, + { + "epoch": 11.458466453674122, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 14346 + }, + { + "epoch": 11.45926517571885, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 14347 + }, + { + "epoch": 11.460063897763579, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 14348 + }, + { + "epoch": 11.460862619808307, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 14349 + }, + { + "epoch": 11.461661341853036, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 14350 + }, + { + "epoch": 11.462460063897764, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 14351 + }, + { + "epoch": 11.463258785942491, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 14352 + }, + { + "epoch": 11.46405750798722, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 14353 + }, + { + "epoch": 11.464856230031948, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 14354 + }, + { + "epoch": 11.465654952076678, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 14355 + }, + { + "epoch": 11.466453674121405, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 14356 + }, + { + "epoch": 11.467252396166135, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 14357 + }, + { + "epoch": 11.468051118210862, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14358 + }, + { + "epoch": 11.468849840255592, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 14359 + }, + { + "epoch": 11.46964856230032, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 14360 + }, + { + "epoch": 11.470447284345047, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14361 + }, + { + "epoch": 11.471246006389777, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 14362 + }, + { + "epoch": 11.472044728434504, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 14363 + }, + { + "epoch": 11.472843450479234, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 14364 + }, + { + "epoch": 11.473642172523961, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 14365 + }, + { + "epoch": 11.47444089456869, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 14366 + }, + { + "epoch": 11.475239616613418, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14367 + }, + { + "epoch": 11.476038338658148, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 14368 + }, + { + "epoch": 11.476837060702875, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 14369 + }, + { + "epoch": 11.477635782747603, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 14370 + }, + { + "epoch": 11.478434504792332, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 14371 + }, + { + "epoch": 11.47923322683706, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 14372 + }, + { + "epoch": 11.48003194888179, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14373 + }, + { + "epoch": 11.480830670926517, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 14374 + }, + { + "epoch": 11.481629392971247, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 14375 + }, + { + "epoch": 11.482428115015974, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 14376 + }, + { + "epoch": 11.483226837060704, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 14377 + }, + { + "epoch": 11.484025559105431, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 14378 + }, + { + "epoch": 11.48482428115016, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 14379 + }, + { + "epoch": 11.485623003194888, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 14380 + }, + { + "epoch": 11.486421725239616, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 14381 + }, + { + "epoch": 11.487220447284345, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 14382 + }, + { + "epoch": 11.488019169329073, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 14383 + }, + { + "epoch": 11.488817891373802, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 14384 + }, + { + "epoch": 11.48961661341853, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 14385 + }, + { + "epoch": 11.49041533546326, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 14386 + }, + { + "epoch": 11.491214057507987, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 14387 + }, + { + "epoch": 11.492012779552716, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14388 + }, + { + "epoch": 11.492811501597444, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 14389 + }, + { + "epoch": 11.493610223642172, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 14390 + }, + { + "epoch": 11.494408945686901, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 14391 + }, + { + "epoch": 11.495207667731629, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 14392 + }, + { + "epoch": 11.496006389776358, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 14393 + }, + { + "epoch": 11.496805111821086, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 14394 + }, + { + "epoch": 11.497603833865815, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 14395 + }, + { + "epoch": 11.498402555910543, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 14396 + }, + { + "epoch": 11.499201277955272, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 14397 + }, + { + "epoch": 11.5, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 14398 + }, + { + "epoch": 11.500798722044728, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 14399 + }, + { + "epoch": 11.501597444089457, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14400 + }, + { + "epoch": 11.502396166134185, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 14401 + }, + { + "epoch": 11.503194888178914, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 14402 + }, + { + "epoch": 11.503993610223642, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 14403 + }, + { + "epoch": 11.504792332268371, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 14404 + }, + { + "epoch": 11.505591054313099, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 14405 + }, + { + "epoch": 11.506389776357828, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 14406 + }, + { + "epoch": 11.507188498402556, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 14407 + }, + { + "epoch": 11.507987220447284, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 14408 + }, + { + "epoch": 11.508785942492013, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 14409 + }, + { + "epoch": 11.50958466453674, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 14410 + }, + { + "epoch": 11.51038338658147, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 14411 + }, + { + "epoch": 11.511182108626198, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 14412 + }, + { + "epoch": 11.511980830670927, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 14413 + }, + { + "epoch": 11.512779552715655, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 14414 + }, + { + "epoch": 11.513578274760384, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 14415 + }, + { + "epoch": 11.514376996805112, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 14416 + }, + { + "epoch": 11.51517571884984, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 14417 + }, + { + "epoch": 11.515974440894569, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 14418 + }, + { + "epoch": 11.516773162939296, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14419 + }, + { + "epoch": 11.517571884984026, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 14420 + }, + { + "epoch": 11.518370607028753, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14421 + }, + { + "epoch": 11.519169329073483, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 14422 + }, + { + "epoch": 11.51996805111821, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 14423 + }, + { + "epoch": 11.52076677316294, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 14424 + }, + { + "epoch": 11.521565495207668, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 14425 + }, + { + "epoch": 11.522364217252395, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 14426 + }, + { + "epoch": 11.523162939297125, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 14427 + }, + { + "epoch": 11.523961661341852, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 14428 + }, + { + "epoch": 11.524760383386582, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 14429 + }, + { + "epoch": 11.52555910543131, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 14430 + }, + { + "epoch": 11.526357827476039, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 14431 + }, + { + "epoch": 11.527156549520766, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 14432 + }, + { + "epoch": 11.527955271565496, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 14433 + }, + { + "epoch": 11.528753993610223, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 14434 + }, + { + "epoch": 11.529552715654953, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 14435 + }, + { + "epoch": 11.53035143769968, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 14436 + }, + { + "epoch": 11.531150159744408, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 14437 + }, + { + "epoch": 11.531948881789138, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1118, + "step": 14438 + }, + { + "epoch": 11.532747603833865, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 14439 + }, + { + "epoch": 11.533546325878595, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 14440 + }, + { + "epoch": 11.534345047923322, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 14441 + }, + { + "epoch": 11.535143769968052, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 14442 + }, + { + "epoch": 11.53594249201278, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 14443 + }, + { + "epoch": 11.536741214057509, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 14444 + }, + { + "epoch": 11.537539936102236, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 14445 + }, + { + "epoch": 11.538338658146966, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 14446 + }, + { + "epoch": 11.539137380191693, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 14447 + }, + { + "epoch": 11.539936102236421, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 14448 + }, + { + "epoch": 11.54073482428115, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 14449 + }, + { + "epoch": 11.541533546325878, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 14450 + }, + { + "epoch": 11.542332268370608, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 14451 + }, + { + "epoch": 11.543130990415335, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 14452 + }, + { + "epoch": 11.543929712460065, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 14453 + }, + { + "epoch": 11.544728434504792, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 14454 + }, + { + "epoch": 11.545527156549522, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 14455 + }, + { + "epoch": 11.54632587859425, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 14456 + }, + { + "epoch": 11.547124600638977, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 14457 + }, + { + "epoch": 11.547923322683706, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 14458 + }, + { + "epoch": 11.548722044728434, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 14459 + }, + { + "epoch": 11.549520766773163, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 14460 + }, + { + "epoch": 11.550319488817891, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 14461 + }, + { + "epoch": 11.55111821086262, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 14462 + }, + { + "epoch": 11.551916932907348, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 14463 + }, + { + "epoch": 11.552715654952078, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1141, + "step": 14464 + }, + { + "epoch": 11.553514376996805, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 14465 + }, + { + "epoch": 11.554313099041533, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 14466 + }, + { + "epoch": 11.555111821086262, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 14467 + }, + { + "epoch": 11.55591054313099, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 14468 + }, + { + "epoch": 11.55670926517572, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 14469 + }, + { + "epoch": 11.557507987220447, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14470 + }, + { + "epoch": 11.558306709265176, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 14471 + }, + { + "epoch": 11.559105431309904, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 14472 + }, + { + "epoch": 11.559904153354633, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 14473 + }, + { + "epoch": 11.560702875399361, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 14474 + }, + { + "epoch": 11.561501597444089, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 14475 + }, + { + "epoch": 11.562300319488818, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 14476 + }, + { + "epoch": 11.563099041533546, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 14477 + }, + { + "epoch": 11.563897763578275, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 14478 + }, + { + "epoch": 11.564696485623003, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 14479 + }, + { + "epoch": 11.565495207667732, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 14480 + }, + { + "epoch": 11.56629392971246, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1154, + "step": 14481 + }, + { + "epoch": 11.56709265175719, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 14482 + }, + { + "epoch": 11.567891373801917, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 14483 + }, + { + "epoch": 11.568690095846645, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 14484 + }, + { + "epoch": 11.569488817891374, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 14485 + }, + { + "epoch": 11.570287539936102, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14486 + }, + { + "epoch": 11.571086261980831, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14487 + }, + { + "epoch": 11.571884984025559, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 14488 + }, + { + "epoch": 11.572683706070288, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14489 + }, + { + "epoch": 11.573482428115016, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 14490 + }, + { + "epoch": 11.574281150159745, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 14491 + }, + { + "epoch": 11.575079872204473, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 14492 + }, + { + "epoch": 11.5758785942492, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 14493 + }, + { + "epoch": 11.57667731629393, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 14494 + }, + { + "epoch": 11.577476038338657, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 14495 + }, + { + "epoch": 11.578274760383387, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 14496 + }, + { + "epoch": 11.579073482428115, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 14497 + }, + { + "epoch": 11.579872204472844, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14498 + }, + { + "epoch": 11.580670926517572, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 14499 + }, + { + "epoch": 11.581469648562301, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 14500 + }, + { + "epoch": 11.582268370607029, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 14501 + }, + { + "epoch": 11.583067092651756, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 14502 + }, + { + "epoch": 11.583865814696486, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 14503 + }, + { + "epoch": 11.584664536741213, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14504 + }, + { + "epoch": 11.585463258785943, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 14505 + }, + { + "epoch": 11.58626198083067, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 14506 + }, + { + "epoch": 11.5870607028754, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 14507 + }, + { + "epoch": 11.587859424920127, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 14508 + }, + { + "epoch": 11.588658146964857, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 14509 + }, + { + "epoch": 11.589456869009584, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 14510 + }, + { + "epoch": 11.590255591054314, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14511 + }, + { + "epoch": 11.591054313099042, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 14512 + }, + { + "epoch": 11.59185303514377, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1124, + "step": 14513 + }, + { + "epoch": 11.592651757188499, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 14514 + }, + { + "epoch": 11.593450479233226, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 14515 + }, + { + "epoch": 11.594249201277956, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 14516 + }, + { + "epoch": 11.595047923322683, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 14517 + }, + { + "epoch": 11.595846645367413, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14518 + }, + { + "epoch": 11.59664536741214, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 14519 + }, + { + "epoch": 11.59744408945687, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 14520 + }, + { + "epoch": 11.598242811501597, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 14521 + }, + { + "epoch": 11.599041533546325, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 14522 + }, + { + "epoch": 11.599840255591054, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 14523 + }, + { + "epoch": 11.600638977635782, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 14524 + }, + { + "epoch": 11.601437699680512, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 14525 + }, + { + "epoch": 11.60223642172524, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 14526 + }, + { + "epoch": 11.603035143769969, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 14527 + }, + { + "epoch": 11.603833865814696, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 14528 + }, + { + "epoch": 11.604632587859426, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 14529 + }, + { + "epoch": 11.605431309904153, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 14530 + }, + { + "epoch": 11.606230031948883, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 14531 + }, + { + "epoch": 11.60702875399361, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 14532 + }, + { + "epoch": 11.607827476038338, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 14533 + }, + { + "epoch": 11.608626198083067, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 14534 + }, + { + "epoch": 11.609424920127795, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 14535 + }, + { + "epoch": 11.610223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 14536 + }, + { + "epoch": 11.611022364217252, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 14537 + }, + { + "epoch": 11.611821086261982, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 14538 + }, + { + "epoch": 11.61261980830671, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 14539 + }, + { + "epoch": 11.613418530351439, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 14540 + }, + { + "epoch": 11.614217252396166, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 14541 + }, + { + "epoch": 11.615015974440894, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 14542 + }, + { + "epoch": 11.615814696485623, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 14543 + }, + { + "epoch": 11.616613418530351, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 14544 + }, + { + "epoch": 11.61741214057508, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 14545 + }, + { + "epoch": 11.618210862619808, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 14546 + }, + { + "epoch": 11.619009584664537, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 14547 + }, + { + "epoch": 11.619808306709265, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 14548 + }, + { + "epoch": 11.620607028753994, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 14549 + }, + { + "epoch": 11.621405750798722, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 14550 + }, + { + "epoch": 11.62220447284345, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 14551 + }, + { + "epoch": 11.62300319488818, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 14552 + }, + { + "epoch": 11.623801916932907, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 14553 + }, + { + "epoch": 11.624600638977636, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 14554 + }, + { + "epoch": 11.625399361022364, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 14555 + }, + { + "epoch": 11.626198083067093, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14556 + }, + { + "epoch": 11.62699680511182, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14557 + }, + { + "epoch": 11.62779552715655, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 14558 + }, + { + "epoch": 11.628594249201278, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 14559 + }, + { + "epoch": 11.629392971246006, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 14560 + }, + { + "epoch": 11.630191693290735, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 14561 + }, + { + "epoch": 11.630990415335463, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 14562 + }, + { + "epoch": 11.631789137380192, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 14563 + }, + { + "epoch": 11.63258785942492, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 14564 + }, + { + "epoch": 11.633386581469649, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 14565 + }, + { + "epoch": 11.634185303514377, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 14566 + }, + { + "epoch": 11.634984025559106, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 14567 + }, + { + "epoch": 11.635782747603834, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 14568 + }, + { + "epoch": 11.636581469648561, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 14569 + }, + { + "epoch": 11.63738019169329, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 14570 + }, + { + "epoch": 11.638178913738018, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 14571 + }, + { + "epoch": 11.638977635782748, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 14572 + }, + { + "epoch": 11.639776357827476, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 14573 + }, + { + "epoch": 11.640575079872205, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 14574 + }, + { + "epoch": 11.641373801916933, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 14575 + }, + { + "epoch": 11.642172523961662, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 14576 + }, + { + "epoch": 11.64297124600639, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 14577 + }, + { + "epoch": 11.643769968051117, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 14578 + }, + { + "epoch": 11.644568690095847, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 14579 + }, + { + "epoch": 11.645367412140574, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 14580 + }, + { + "epoch": 11.646166134185304, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 14581 + }, + { + "epoch": 11.646964856230031, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 14582 + }, + { + "epoch": 11.64776357827476, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 14583 + }, + { + "epoch": 11.648562300319488, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 14584 + }, + { + "epoch": 11.649361022364218, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 14585 + }, + { + "epoch": 11.650159744408946, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 14586 + }, + { + "epoch": 11.650958466453675, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14587 + }, + { + "epoch": 11.651757188498403, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 14588 + }, + { + "epoch": 11.65255591054313, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1149, + "step": 14589 + }, + { + "epoch": 11.65335463258786, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 14590 + }, + { + "epoch": 11.654153354632587, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 14591 + }, + { + "epoch": 11.654952076677317, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 14592 + }, + { + "epoch": 11.655750798722044, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 14593 + }, + { + "epoch": 11.656549520766774, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 14594 + }, + { + "epoch": 11.657348242811501, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 14595 + }, + { + "epoch": 11.65814696485623, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 14596 + }, + { + "epoch": 11.658945686900958, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 14597 + }, + { + "epoch": 11.659744408945686, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 14598 + }, + { + "epoch": 11.660543130990416, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 14599 + }, + { + "epoch": 11.661341853035143, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 14600 + }, + { + "epoch": 11.662140575079873, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 14601 + }, + { + "epoch": 11.6629392971246, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 14602 + }, + { + "epoch": 11.66373801916933, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14603 + }, + { + "epoch": 11.664536741214057, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14604 + }, + { + "epoch": 11.665335463258787, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 14605 + }, + { + "epoch": 11.666134185303514, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 14606 + }, + { + "epoch": 11.666932907348244, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 14607 + }, + { + "epoch": 11.667731629392971, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 14608 + }, + { + "epoch": 11.668530351437699, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 14609 + }, + { + "epoch": 11.669329073482428, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 14610 + }, + { + "epoch": 11.670127795527156, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 14611 + }, + { + "epoch": 11.670926517571885, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 14612 + }, + { + "epoch": 11.671725239616613, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 14613 + }, + { + "epoch": 11.672523961661343, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 14614 + }, + { + "epoch": 11.67332268370607, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 14615 + }, + { + "epoch": 11.6741214057508, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 14616 + }, + { + "epoch": 11.674920127795527, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 14617 + }, + { + "epoch": 11.675718849840255, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14618 + }, + { + "epoch": 11.676517571884984, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 14619 + }, + { + "epoch": 11.677316293929712, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 14620 + }, + { + "epoch": 11.678115015974441, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 14621 + }, + { + "epoch": 11.678913738019169, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 14622 + }, + { + "epoch": 11.679712460063898, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14623 + }, + { + "epoch": 11.680511182108626, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 14624 + }, + { + "epoch": 11.681309904153355, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 14625 + }, + { + "epoch": 11.682108626198083, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 14626 + }, + { + "epoch": 11.68290734824281, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 14627 + }, + { + "epoch": 11.68370607028754, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 14628 + }, + { + "epoch": 11.684504792332268, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 14629 + }, + { + "epoch": 11.685303514376997, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 14630 + }, + { + "epoch": 11.686102236421725, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 14631 + }, + { + "epoch": 11.686900958466454, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 14632 + }, + { + "epoch": 11.687699680511182, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 14633 + }, + { + "epoch": 11.688498402555911, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 14634 + }, + { + "epoch": 11.689297124600639, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 14635 + }, + { + "epoch": 11.690095846645367, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 14636 + }, + { + "epoch": 11.690894568690096, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 14637 + }, + { + "epoch": 11.691693290734824, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14638 + }, + { + "epoch": 11.692492012779553, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 14639 + }, + { + "epoch": 11.69329073482428, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 14640 + }, + { + "epoch": 11.69408945686901, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1159, + "step": 14641 + }, + { + "epoch": 11.694888178913738, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 14642 + }, + { + "epoch": 11.695686900958467, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 14643 + }, + { + "epoch": 11.696485623003195, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 14644 + }, + { + "epoch": 11.697284345047922, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 14645 + }, + { + "epoch": 11.698083067092652, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 14646 + }, + { + "epoch": 11.69888178913738, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 14647 + }, + { + "epoch": 11.699680511182109, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 14648 + }, + { + "epoch": 11.700479233226837, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 14649 + }, + { + "epoch": 11.701277955271566, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 14650 + }, + { + "epoch": 11.702076677316294, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 14651 + }, + { + "epoch": 11.702875399361023, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 14652 + }, + { + "epoch": 11.70367412140575, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 14653 + }, + { + "epoch": 11.704472843450478, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 14654 + }, + { + "epoch": 11.705271565495208, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 14655 + }, + { + "epoch": 11.706070287539935, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 14656 + }, + { + "epoch": 11.706869009584665, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 14657 + }, + { + "epoch": 11.707667731629392, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 14658 + }, + { + "epoch": 11.708466453674122, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 14659 + }, + { + "epoch": 11.70926517571885, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 14660 + }, + { + "epoch": 11.710063897763579, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14661 + }, + { + "epoch": 11.710862619808307, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 14662 + }, + { + "epoch": 11.711661341853034, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 14663 + }, + { + "epoch": 11.712460063897764, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 14664 + }, + { + "epoch": 11.713258785942491, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 14665 + }, + { + "epoch": 11.71405750798722, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 14666 + }, + { + "epoch": 11.714856230031948, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 14667 + }, + { + "epoch": 11.715654952076678, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 14668 + }, + { + "epoch": 11.716453674121405, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 14669 + }, + { + "epoch": 11.717252396166135, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 14670 + }, + { + "epoch": 11.718051118210862, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 14671 + }, + { + "epoch": 11.718849840255592, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 14672 + }, + { + "epoch": 11.71964856230032, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 14673 + }, + { + "epoch": 11.720447284345047, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 14674 + }, + { + "epoch": 11.721246006389777, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 14675 + }, + { + "epoch": 11.722044728434504, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 14676 + }, + { + "epoch": 11.722843450479234, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 14677 + }, + { + "epoch": 11.723642172523961, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14678 + }, + { + "epoch": 11.72444089456869, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 14679 + }, + { + "epoch": 11.725239616613418, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 14680 + }, + { + "epoch": 11.726038338658148, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 14681 + }, + { + "epoch": 11.726837060702875, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 14682 + }, + { + "epoch": 11.727635782747605, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14683 + }, + { + "epoch": 11.728434504792332, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 14684 + }, + { + "epoch": 11.72923322683706, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 14685 + }, + { + "epoch": 11.73003194888179, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 14686 + }, + { + "epoch": 11.730830670926517, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 14687 + }, + { + "epoch": 11.731629392971247, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 14688 + }, + { + "epoch": 11.732428115015974, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 14689 + }, + { + "epoch": 11.733226837060704, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 14690 + }, + { + "epoch": 11.734025559105431, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 14691 + }, + { + "epoch": 11.73482428115016, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 14692 + }, + { + "epoch": 11.735623003194888, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14693 + }, + { + "epoch": 11.736421725239616, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 14694 + }, + { + "epoch": 11.737220447284345, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 14695 + }, + { + "epoch": 11.738019169329073, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 14696 + }, + { + "epoch": 11.738817891373802, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 14697 + }, + { + "epoch": 11.73961661341853, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 14698 + }, + { + "epoch": 11.74041533546326, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 14699 + }, + { + "epoch": 11.741214057507987, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 14700 + }, + { + "epoch": 11.742012779552716, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 14701 + }, + { + "epoch": 11.742811501597444, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 14702 + }, + { + "epoch": 11.743610223642172, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 14703 + }, + { + "epoch": 11.744408945686901, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 14704 + }, + { + "epoch": 11.745207667731629, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 14705 + }, + { + "epoch": 11.746006389776358, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 14706 + }, + { + "epoch": 11.746805111821086, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14707 + }, + { + "epoch": 11.747603833865815, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 14708 + }, + { + "epoch": 11.748402555910543, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 14709 + }, + { + "epoch": 11.749201277955272, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 14710 + }, + { + "epoch": 11.75, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 14711 + }, + { + "epoch": 11.750798722044728, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 14712 + }, + { + "epoch": 11.751597444089457, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 14713 + }, + { + "epoch": 11.752396166134185, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 14714 + }, + { + "epoch": 11.753194888178914, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 14715 + }, + { + "epoch": 11.753993610223642, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 14716 + }, + { + "epoch": 11.754792332268371, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 14717 + }, + { + "epoch": 11.755591054313099, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 14718 + }, + { + "epoch": 11.756389776357828, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 14719 + }, + { + "epoch": 11.757188498402556, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 14720 + }, + { + "epoch": 11.757987220447284, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 14721 + }, + { + "epoch": 11.758785942492013, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 14722 + }, + { + "epoch": 11.75958466453674, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 14723 + }, + { + "epoch": 11.76038338658147, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 14724 + }, + { + "epoch": 11.761182108626198, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 14725 + }, + { + "epoch": 11.761980830670927, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 14726 + }, + { + "epoch": 11.762779552715655, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 14727 + }, + { + "epoch": 11.763578274760384, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 14728 + }, + { + "epoch": 11.764376996805112, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 14729 + }, + { + "epoch": 11.76517571884984, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 14730 + }, + { + "epoch": 11.765974440894569, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 14731 + }, + { + "epoch": 11.766773162939296, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 14732 + }, + { + "epoch": 11.767571884984026, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 14733 + }, + { + "epoch": 11.768370607028753, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 14734 + }, + { + "epoch": 11.769169329073483, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 14735 + }, + { + "epoch": 11.76996805111821, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 14736 + }, + { + "epoch": 11.77076677316294, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 14737 + }, + { + "epoch": 11.771565495207668, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 14738 + }, + { + "epoch": 11.772364217252395, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 14739 + }, + { + "epoch": 11.773162939297125, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14740 + }, + { + "epoch": 11.773961661341852, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 14741 + }, + { + "epoch": 11.774760383386582, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 14742 + }, + { + "epoch": 11.77555910543131, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 14743 + }, + { + "epoch": 11.776357827476039, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1131, + "step": 14744 + }, + { + "epoch": 11.777156549520766, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 14745 + }, + { + "epoch": 11.777955271565496, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 14746 + }, + { + "epoch": 11.778753993610223, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 14747 + }, + { + "epoch": 11.779552715654953, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 14748 + }, + { + "epoch": 11.78035143769968, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 14749 + }, + { + "epoch": 11.781150159744408, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 14750 + }, + { + "epoch": 11.781948881789138, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 14751 + }, + { + "epoch": 11.782747603833865, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 14752 + }, + { + "epoch": 11.783546325878595, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 14753 + }, + { + "epoch": 11.784345047923322, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 14754 + }, + { + "epoch": 11.785143769968052, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 14755 + }, + { + "epoch": 11.78594249201278, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 14756 + }, + { + "epoch": 11.786741214057509, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 14757 + }, + { + "epoch": 11.787539936102236, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 14758 + }, + { + "epoch": 11.788338658146966, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 14759 + }, + { + "epoch": 11.789137380191693, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 14760 + }, + { + "epoch": 11.789936102236421, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 14761 + }, + { + "epoch": 11.79073482428115, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 14762 + }, + { + "epoch": 11.791533546325878, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 14763 + }, + { + "epoch": 11.792332268370608, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 14764 + }, + { + "epoch": 11.793130990415335, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 14765 + }, + { + "epoch": 11.793929712460065, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 14766 + }, + { + "epoch": 11.794728434504792, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 14767 + }, + { + "epoch": 11.795527156549522, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 14768 + }, + { + "epoch": 11.79632587859425, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 14769 + }, + { + "epoch": 11.797124600638977, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 14770 + }, + { + "epoch": 11.797923322683706, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 14771 + }, + { + "epoch": 11.798722044728434, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 14772 + }, + { + "epoch": 11.799520766773163, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 14773 + }, + { + "epoch": 11.800319488817891, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 14774 + }, + { + "epoch": 11.80111821086262, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14775 + }, + { + "epoch": 11.801916932907348, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 14776 + }, + { + "epoch": 11.802715654952078, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1121, + "step": 14777 + }, + { + "epoch": 11.803514376996805, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 14778 + }, + { + "epoch": 11.804313099041533, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 14779 + }, + { + "epoch": 11.805111821086262, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 14780 + }, + { + "epoch": 11.80591054313099, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14781 + }, + { + "epoch": 11.80670926517572, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14782 + }, + { + "epoch": 11.807507987220447, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14783 + }, + { + "epoch": 11.808306709265176, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 14784 + }, + { + "epoch": 11.809105431309904, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 14785 + }, + { + "epoch": 11.809904153354633, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 14786 + }, + { + "epoch": 11.810702875399361, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 14787 + }, + { + "epoch": 11.811501597444089, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 14788 + }, + { + "epoch": 11.812300319488818, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 14789 + }, + { + "epoch": 11.813099041533546, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14790 + }, + { + "epoch": 11.813897763578275, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 14791 + }, + { + "epoch": 11.814696485623003, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 14792 + }, + { + "epoch": 11.815495207667732, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 14793 + }, + { + "epoch": 11.81629392971246, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 14794 + }, + { + "epoch": 11.81709265175719, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 14795 + }, + { + "epoch": 11.817891373801917, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 14796 + }, + { + "epoch": 11.818690095846645, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 14797 + }, + { + "epoch": 11.819488817891374, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 14798 + }, + { + "epoch": 11.820287539936102, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14799 + }, + { + "epoch": 11.821086261980831, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 14800 + }, + { + "epoch": 11.821884984025559, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 14801 + }, + { + "epoch": 11.822683706070288, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 14802 + }, + { + "epoch": 11.823482428115016, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 14803 + }, + { + "epoch": 11.824281150159745, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 14804 + }, + { + "epoch": 11.825079872204473, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 14805 + }, + { + "epoch": 11.8258785942492, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 14806 + }, + { + "epoch": 11.82667731629393, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 14807 + }, + { + "epoch": 11.827476038338657, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 14808 + }, + { + "epoch": 11.828274760383387, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 14809 + }, + { + "epoch": 11.829073482428115, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 14810 + }, + { + "epoch": 11.829872204472844, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 14811 + }, + { + "epoch": 11.830670926517572, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 14812 + }, + { + "epoch": 11.831469648562301, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 14813 + }, + { + "epoch": 11.832268370607029, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 14814 + }, + { + "epoch": 11.833067092651756, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 14815 + }, + { + "epoch": 11.833865814696486, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 14816 + }, + { + "epoch": 11.834664536741213, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 14817 + }, + { + "epoch": 11.835463258785943, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 14818 + }, + { + "epoch": 11.83626198083067, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 14819 + }, + { + "epoch": 11.8370607028754, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 14820 + }, + { + "epoch": 11.837859424920127, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 14821 + }, + { + "epoch": 11.838658146964857, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 14822 + }, + { + "epoch": 11.839456869009584, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 14823 + }, + { + "epoch": 11.840255591054314, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14824 + }, + { + "epoch": 11.841054313099042, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 14825 + }, + { + "epoch": 11.84185303514377, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 14826 + }, + { + "epoch": 11.842651757188499, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14827 + }, + { + "epoch": 11.843450479233226, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1109, + "step": 14828 + }, + { + "epoch": 11.844249201277956, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 14829 + }, + { + "epoch": 11.845047923322683, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 14830 + }, + { + "epoch": 11.845846645367413, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 14831 + }, + { + "epoch": 11.84664536741214, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 14832 + }, + { + "epoch": 11.84744408945687, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 14833 + }, + { + "epoch": 11.848242811501597, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 14834 + }, + { + "epoch": 11.849041533546325, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 14835 + }, + { + "epoch": 11.849840255591054, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 14836 + }, + { + "epoch": 11.850638977635782, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 14837 + }, + { + "epoch": 11.851437699680512, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 14838 + }, + { + "epoch": 11.85223642172524, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 14839 + }, + { + "epoch": 11.853035143769969, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 14840 + }, + { + "epoch": 11.853833865814696, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 14841 + }, + { + "epoch": 11.854632587859426, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 14842 + }, + { + "epoch": 11.855431309904153, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 14843 + }, + { + "epoch": 11.856230031948883, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 14844 + }, + { + "epoch": 11.85702875399361, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 14845 + }, + { + "epoch": 11.857827476038338, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 14846 + }, + { + "epoch": 11.858626198083067, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14847 + }, + { + "epoch": 11.859424920127795, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 14848 + }, + { + "epoch": 11.860223642172524, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 14849 + }, + { + "epoch": 11.861022364217252, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 14850 + }, + { + "epoch": 11.861821086261982, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 14851 + }, + { + "epoch": 11.86261980830671, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 14852 + }, + { + "epoch": 11.863418530351439, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 14853 + }, + { + "epoch": 11.864217252396166, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 14854 + }, + { + "epoch": 11.865015974440894, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 14855 + }, + { + "epoch": 11.865814696485623, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 14856 + }, + { + "epoch": 11.866613418530351, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 14857 + }, + { + "epoch": 11.86741214057508, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 14858 + }, + { + "epoch": 11.868210862619808, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 14859 + }, + { + "epoch": 11.869009584664537, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 14860 + }, + { + "epoch": 11.869808306709265, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14861 + }, + { + "epoch": 11.870607028753994, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 14862 + }, + { + "epoch": 11.871405750798722, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 14863 + }, + { + "epoch": 11.87220447284345, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14864 + }, + { + "epoch": 11.87300319488818, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 14865 + }, + { + "epoch": 11.873801916932907, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 14866 + }, + { + "epoch": 11.874600638977636, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 14867 + }, + { + "epoch": 11.875399361022364, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14868 + }, + { + "epoch": 11.876198083067093, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 14869 + }, + { + "epoch": 11.87699680511182, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 14870 + }, + { + "epoch": 11.87779552715655, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 14871 + }, + { + "epoch": 11.878594249201278, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 14872 + }, + { + "epoch": 11.879392971246006, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 14873 + }, + { + "epoch": 11.880191693290735, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 14874 + }, + { + "epoch": 11.880990415335463, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 14875 + }, + { + "epoch": 11.881789137380192, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 14876 + }, + { + "epoch": 11.88258785942492, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 14877 + }, + { + "epoch": 11.883386581469649, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 14878 + }, + { + "epoch": 11.884185303514377, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 14879 + }, + { + "epoch": 11.884984025559106, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 14880 + }, + { + "epoch": 11.885782747603834, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14881 + }, + { + "epoch": 11.886581469648561, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 14882 + }, + { + "epoch": 11.88738019169329, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 14883 + }, + { + "epoch": 11.888178913738018, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 14884 + }, + { + "epoch": 11.888977635782748, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 14885 + }, + { + "epoch": 11.889776357827476, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14886 + }, + { + "epoch": 11.890575079872205, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 14887 + }, + { + "epoch": 11.891373801916933, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 14888 + }, + { + "epoch": 11.892172523961662, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 14889 + }, + { + "epoch": 11.89297124600639, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 14890 + }, + { + "epoch": 11.893769968051117, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14891 + }, + { + "epoch": 11.894568690095847, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 14892 + }, + { + "epoch": 11.895367412140574, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 14893 + }, + { + "epoch": 11.896166134185304, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 14894 + }, + { + "epoch": 11.896964856230031, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 14895 + }, + { + "epoch": 11.89776357827476, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 14896 + }, + { + "epoch": 11.898562300319488, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 14897 + }, + { + "epoch": 11.899361022364218, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14898 + }, + { + "epoch": 11.900159744408946, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 14899 + }, + { + "epoch": 11.900958466453675, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 14900 + }, + { + "epoch": 11.901757188498403, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 14901 + }, + { + "epoch": 11.90255591054313, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 14902 + }, + { + "epoch": 11.90335463258786, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 14903 + }, + { + "epoch": 11.904153354632587, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 14904 + }, + { + "epoch": 11.904952076677317, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 14905 + }, + { + "epoch": 11.905750798722044, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 14906 + }, + { + "epoch": 11.906549520766774, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 14907 + }, + { + "epoch": 11.907348242811501, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 14908 + }, + { + "epoch": 11.90814696485623, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 14909 + }, + { + "epoch": 11.908945686900958, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 14910 + }, + { + "epoch": 11.909744408945686, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 14911 + }, + { + "epoch": 11.910543130990416, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 14912 + }, + { + "epoch": 11.911341853035143, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 14913 + }, + { + "epoch": 11.912140575079873, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 14914 + }, + { + "epoch": 11.9129392971246, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 14915 + }, + { + "epoch": 11.91373801916933, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 14916 + }, + { + "epoch": 11.914536741214057, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14917 + }, + { + "epoch": 11.915335463258787, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 14918 + }, + { + "epoch": 11.916134185303514, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 14919 + }, + { + "epoch": 11.916932907348244, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 14920 + }, + { + "epoch": 11.917731629392971, + "grad_norm": 0.09423828125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 14921 + }, + { + "epoch": 11.918530351437699, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 14922 + }, + { + "epoch": 11.919329073482428, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 14923 + }, + { + "epoch": 11.920127795527156, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 14924 + }, + { + "epoch": 11.920926517571885, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 14925 + }, + { + "epoch": 11.921725239616613, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 14926 + }, + { + "epoch": 11.922523961661343, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14927 + }, + { + "epoch": 11.92332268370607, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 14928 + }, + { + "epoch": 11.9241214057508, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 14929 + }, + { + "epoch": 11.924920127795527, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 14930 + }, + { + "epoch": 11.925718849840255, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 14931 + }, + { + "epoch": 11.926517571884984, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 14932 + }, + { + "epoch": 11.927316293929712, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 14933 + }, + { + "epoch": 11.928115015974441, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 14934 + }, + { + "epoch": 11.928913738019169, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 14935 + }, + { + "epoch": 11.929712460063898, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14936 + }, + { + "epoch": 11.930511182108626, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 14937 + }, + { + "epoch": 11.931309904153355, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 14938 + }, + { + "epoch": 11.932108626198083, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 14939 + }, + { + "epoch": 11.93290734824281, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 14940 + }, + { + "epoch": 11.93370607028754, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 14941 + }, + { + "epoch": 11.934504792332268, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 14942 + }, + { + "epoch": 11.935303514376997, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 14943 + }, + { + "epoch": 11.936102236421725, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 14944 + }, + { + "epoch": 11.936900958466454, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 14945 + }, + { + "epoch": 11.937699680511182, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 14946 + }, + { + "epoch": 11.938498402555911, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 14947 + }, + { + "epoch": 11.939297124600639, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 14948 + }, + { + "epoch": 11.940095846645367, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 14949 + }, + { + "epoch": 11.940894568690096, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 14950 + }, + { + "epoch": 11.941693290734824, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14951 + }, + { + "epoch": 11.942492012779553, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 14952 + }, + { + "epoch": 11.94329073482428, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 14953 + }, + { + "epoch": 11.94408945686901, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 14954 + }, + { + "epoch": 11.944888178913738, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 14955 + }, + { + "epoch": 11.945686900958467, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 14956 + }, + { + "epoch": 11.946485623003195, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 14957 + }, + { + "epoch": 11.947284345047922, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 14958 + }, + { + "epoch": 11.948083067092652, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 14959 + }, + { + "epoch": 11.94888178913738, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 14960 + }, + { + "epoch": 11.949680511182109, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 14961 + }, + { + "epoch": 11.950479233226837, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 14962 + }, + { + "epoch": 11.951277955271566, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 14963 + }, + { + "epoch": 11.952076677316294, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 14964 + }, + { + "epoch": 11.952875399361023, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 14965 + }, + { + "epoch": 11.95367412140575, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 14966 + }, + { + "epoch": 11.954472843450478, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 14967 + }, + { + "epoch": 11.955271565495208, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 14968 + }, + { + "epoch": 11.956070287539935, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 14969 + }, + { + "epoch": 11.956869009584665, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 14970 + }, + { + "epoch": 11.957667731629392, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 14971 + }, + { + "epoch": 11.958466453674122, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 14972 + }, + { + "epoch": 11.95926517571885, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 14973 + }, + { + "epoch": 11.960063897763579, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 14974 + }, + { + "epoch": 11.960862619808307, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 14975 + }, + { + "epoch": 11.961661341853034, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 14976 + }, + { + "epoch": 11.962460063897764, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 14977 + }, + { + "epoch": 11.963258785942491, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 14978 + }, + { + "epoch": 11.96405750798722, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 14979 + }, + { + "epoch": 11.964856230031948, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 14980 + }, + { + "epoch": 11.965654952076678, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 14981 + }, + { + "epoch": 11.966453674121405, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1111, + "step": 14982 + }, + { + "epoch": 11.967252396166135, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 14983 + }, + { + "epoch": 11.968051118210862, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 14984 + }, + { + "epoch": 11.968849840255592, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 14985 + }, + { + "epoch": 11.96964856230032, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 14986 + }, + { + "epoch": 11.970447284345047, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 14987 + }, + { + "epoch": 11.971246006389777, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 14988 + }, + { + "epoch": 11.972044728434504, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 14989 + }, + { + "epoch": 11.972843450479234, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 14990 + }, + { + "epoch": 11.973642172523961, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 14991 + }, + { + "epoch": 11.97444089456869, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 14992 + }, + { + "epoch": 11.975239616613418, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 14993 + }, + { + "epoch": 11.976038338658148, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 14994 + }, + { + "epoch": 11.976837060702875, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 14995 + }, + { + "epoch": 11.977635782747605, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 14996 + }, + { + "epoch": 11.978434504792332, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 14997 + }, + { + "epoch": 11.97923322683706, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 14998 + }, + { + "epoch": 11.98003194888179, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 14999 + }, + { + "epoch": 11.980830670926517, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 15000 + }, + { + "epoch": 11.981629392971247, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 15001 + }, + { + "epoch": 11.982428115015974, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 15002 + }, + { + "epoch": 11.983226837060704, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 15003 + }, + { + "epoch": 11.984025559105431, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 15004 + }, + { + "epoch": 11.98482428115016, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 15005 + }, + { + "epoch": 11.985623003194888, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 15006 + }, + { + "epoch": 11.986421725239616, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 15007 + }, + { + "epoch": 11.987220447284345, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 15008 + }, + { + "epoch": 11.988019169329073, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 15009 + }, + { + "epoch": 11.988817891373802, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 15010 + }, + { + "epoch": 11.98961661341853, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 15011 + }, + { + "epoch": 11.99041533546326, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 15012 + }, + { + "epoch": 11.991214057507987, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 15013 + }, + { + "epoch": 11.992012779552716, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 15014 + }, + { + "epoch": 11.992811501597444, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 15015 + }, + { + "epoch": 11.993610223642172, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 15016 + }, + { + "epoch": 11.994408945686901, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 15017 + }, + { + "epoch": 11.995207667731629, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 15018 + }, + { + "epoch": 11.996006389776358, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 15019 + }, + { + "epoch": 11.996805111821086, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 15020 + }, + { + "epoch": 11.997603833865815, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 15021 + }, + { + "epoch": 11.998402555910543, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 15022 + }, + { + "epoch": 11.999201277955272, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 15023 + }, + { + "epoch": 12.0, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 15024 + }, + { + "epoch": 12.000798722044728, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 15025 + }, + { + "epoch": 12.001597444089457, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 15026 + }, + { + "epoch": 12.002396166134185, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 15027 + }, + { + "epoch": 12.003194888178914, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 15028 + }, + { + "epoch": 12.003993610223642, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 15029 + }, + { + "epoch": 12.004792332268371, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 15030 + }, + { + "epoch": 12.005591054313099, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15031 + }, + { + "epoch": 12.006389776357828, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 15032 + }, + { + "epoch": 12.007188498402556, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 15033 + }, + { + "epoch": 12.007987220447284, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 15034 + }, + { + "epoch": 12.008785942492013, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 15035 + }, + { + "epoch": 12.00958466453674, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 15036 + }, + { + "epoch": 12.01038338658147, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 15037 + }, + { + "epoch": 12.011182108626198, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 15038 + }, + { + "epoch": 12.011980830670927, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 15039 + }, + { + "epoch": 12.012779552715655, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 15040 + }, + { + "epoch": 12.013578274760384, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 15041 + }, + { + "epoch": 12.014376996805112, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 15042 + }, + { + "epoch": 12.01517571884984, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 15043 + }, + { + "epoch": 12.015974440894569, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 15044 + }, + { + "epoch": 12.016773162939296, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 15045 + }, + { + "epoch": 12.017571884984026, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 15046 + }, + { + "epoch": 12.018370607028753, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 15047 + }, + { + "epoch": 12.019169329073483, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 15048 + }, + { + "epoch": 12.01996805111821, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 15049 + }, + { + "epoch": 12.02076677316294, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 15050 + }, + { + "epoch": 12.021565495207668, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15051 + }, + { + "epoch": 12.022364217252397, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 15052 + }, + { + "epoch": 12.023162939297125, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 15053 + }, + { + "epoch": 12.023961661341852, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 15054 + }, + { + "epoch": 12.024760383386582, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 15055 + }, + { + "epoch": 12.02555910543131, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 15056 + }, + { + "epoch": 12.026357827476039, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 15057 + }, + { + "epoch": 12.027156549520766, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 15058 + }, + { + "epoch": 12.027955271565496, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 15059 + }, + { + "epoch": 12.028753993610223, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 15060 + }, + { + "epoch": 12.029552715654953, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 15061 + }, + { + "epoch": 12.03035143769968, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 15062 + }, + { + "epoch": 12.031150159744408, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 15063 + }, + { + "epoch": 12.031948881789138, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 15064 + }, + { + "epoch": 12.032747603833865, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 15065 + }, + { + "epoch": 12.033546325878595, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 15066 + }, + { + "epoch": 12.034345047923322, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 15067 + }, + { + "epoch": 12.035143769968052, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 15068 + }, + { + "epoch": 12.03594249201278, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 15069 + }, + { + "epoch": 12.036741214057509, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 15070 + }, + { + "epoch": 12.037539936102236, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 15071 + }, + { + "epoch": 12.038338658146964, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 15072 + }, + { + "epoch": 12.039137380191693, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 15073 + }, + { + "epoch": 12.039936102236421, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 15074 + }, + { + "epoch": 12.04073482428115, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 15075 + }, + { + "epoch": 12.041533546325878, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 15076 + }, + { + "epoch": 12.042332268370608, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 15077 + }, + { + "epoch": 12.043130990415335, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 15078 + }, + { + "epoch": 12.043929712460065, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 15079 + }, + { + "epoch": 12.044728434504792, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 15080 + }, + { + "epoch": 12.04552715654952, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 15081 + }, + { + "epoch": 12.04632587859425, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 15082 + }, + { + "epoch": 12.047124600638977, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 15083 + }, + { + "epoch": 12.047923322683706, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 15084 + }, + { + "epoch": 12.048722044728434, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 15085 + }, + { + "epoch": 12.049520766773163, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 15086 + }, + { + "epoch": 12.050319488817891, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 15087 + }, + { + "epoch": 12.05111821086262, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 15088 + }, + { + "epoch": 12.051916932907348, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 15089 + }, + { + "epoch": 12.052715654952078, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 15090 + }, + { + "epoch": 12.053514376996805, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 15091 + }, + { + "epoch": 12.054313099041533, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 15092 + }, + { + "epoch": 12.055111821086262, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 15093 + }, + { + "epoch": 12.05591054313099, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 15094 + }, + { + "epoch": 12.05670926517572, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 15095 + }, + { + "epoch": 12.057507987220447, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 15096 + }, + { + "epoch": 12.058306709265176, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 15097 + }, + { + "epoch": 12.059105431309904, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 15098 + }, + { + "epoch": 12.059904153354633, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15099 + }, + { + "epoch": 12.060702875399361, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 15100 + }, + { + "epoch": 12.061501597444089, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 15101 + }, + { + "epoch": 12.062300319488818, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 15102 + }, + { + "epoch": 12.063099041533546, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 15103 + }, + { + "epoch": 12.063897763578275, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 15104 + }, + { + "epoch": 12.064696485623003, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 15105 + }, + { + "epoch": 12.065495207667732, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15106 + }, + { + "epoch": 12.06629392971246, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 15107 + }, + { + "epoch": 12.06709265175719, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 15108 + }, + { + "epoch": 12.067891373801917, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 15109 + }, + { + "epoch": 12.068690095846645, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 15110 + }, + { + "epoch": 12.069488817891374, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 15111 + }, + { + "epoch": 12.070287539936102, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 15112 + }, + { + "epoch": 12.071086261980831, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 15113 + }, + { + "epoch": 12.071884984025559, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 15114 + }, + { + "epoch": 12.072683706070288, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15115 + }, + { + "epoch": 12.073482428115016, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 15116 + }, + { + "epoch": 12.074281150159745, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 15117 + }, + { + "epoch": 12.075079872204473, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 15118 + }, + { + "epoch": 12.0758785942492, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 15119 + }, + { + "epoch": 12.07667731629393, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 15120 + }, + { + "epoch": 12.077476038338657, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 15121 + }, + { + "epoch": 12.078274760383387, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 15122 + }, + { + "epoch": 12.079073482428115, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 15123 + }, + { + "epoch": 12.079872204472844, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 15124 + }, + { + "epoch": 12.080670926517572, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 15125 + }, + { + "epoch": 12.081469648562301, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 15126 + }, + { + "epoch": 12.082268370607029, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 15127 + }, + { + "epoch": 12.083067092651758, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 15128 + }, + { + "epoch": 12.083865814696486, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 15129 + }, + { + "epoch": 12.084664536741213, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 15130 + }, + { + "epoch": 12.085463258785943, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 15131 + }, + { + "epoch": 12.08626198083067, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 15132 + }, + { + "epoch": 12.0870607028754, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 15133 + }, + { + "epoch": 12.087859424920127, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 15134 + }, + { + "epoch": 12.088658146964857, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 15135 + }, + { + "epoch": 12.089456869009584, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 15136 + }, + { + "epoch": 12.090255591054314, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 15137 + }, + { + "epoch": 12.091054313099042, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 15138 + }, + { + "epoch": 12.09185303514377, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 15139 + }, + { + "epoch": 12.092651757188499, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 15140 + }, + { + "epoch": 12.093450479233226, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 15141 + }, + { + "epoch": 12.094249201277956, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 15142 + }, + { + "epoch": 12.095047923322683, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 15143 + }, + { + "epoch": 12.095846645367413, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 15144 + }, + { + "epoch": 12.09664536741214, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 15145 + }, + { + "epoch": 12.09744408945687, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 15146 + }, + { + "epoch": 12.098242811501597, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15147 + }, + { + "epoch": 12.099041533546325, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 15148 + }, + { + "epoch": 12.099840255591054, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 15149 + }, + { + "epoch": 12.100638977635782, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 15150 + }, + { + "epoch": 12.101437699680512, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 15151 + }, + { + "epoch": 12.10223642172524, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 15152 + }, + { + "epoch": 12.103035143769969, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 15153 + }, + { + "epoch": 12.103833865814696, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 15154 + }, + { + "epoch": 12.104632587859426, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 15155 + }, + { + "epoch": 12.105431309904153, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 15156 + }, + { + "epoch": 12.106230031948881, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 15157 + }, + { + "epoch": 12.10702875399361, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 15158 + }, + { + "epoch": 12.107827476038338, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 15159 + }, + { + "epoch": 12.108626198083067, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 15160 + }, + { + "epoch": 12.109424920127795, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 15161 + }, + { + "epoch": 12.110223642172524, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 15162 + }, + { + "epoch": 12.111022364217252, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 15163 + }, + { + "epoch": 12.111821086261982, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 15164 + }, + { + "epoch": 12.11261980830671, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15165 + }, + { + "epoch": 12.113418530351439, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 15166 + }, + { + "epoch": 12.114217252396166, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 15167 + }, + { + "epoch": 12.115015974440894, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15168 + }, + { + "epoch": 12.115814696485623, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 15169 + }, + { + "epoch": 12.116613418530351, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 15170 + }, + { + "epoch": 12.11741214057508, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 15171 + }, + { + "epoch": 12.118210862619808, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 15172 + }, + { + "epoch": 12.119009584664537, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 15173 + }, + { + "epoch": 12.119808306709265, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 15174 + }, + { + "epoch": 12.120607028753994, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 15175 + }, + { + "epoch": 12.121405750798722, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 15176 + }, + { + "epoch": 12.12220447284345, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 15177 + }, + { + "epoch": 12.12300319488818, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 15178 + }, + { + "epoch": 12.123801916932907, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 15179 + }, + { + "epoch": 12.124600638977636, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15180 + }, + { + "epoch": 12.125399361022364, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 15181 + }, + { + "epoch": 12.126198083067093, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 15182 + }, + { + "epoch": 12.12699680511182, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 15183 + }, + { + "epoch": 12.12779552715655, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 15184 + }, + { + "epoch": 12.128594249201278, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 15185 + }, + { + "epoch": 12.129392971246006, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 15186 + }, + { + "epoch": 12.130191693290735, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 15187 + }, + { + "epoch": 12.130990415335463, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 15188 + }, + { + "epoch": 12.131789137380192, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 15189 + }, + { + "epoch": 12.13258785942492, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 15190 + }, + { + "epoch": 12.133386581469649, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 15191 + }, + { + "epoch": 12.134185303514377, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 15192 + }, + { + "epoch": 12.134984025559106, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 15193 + }, + { + "epoch": 12.135782747603834, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 15194 + }, + { + "epoch": 12.136581469648561, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 15195 + }, + { + "epoch": 12.13738019169329, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 15196 + }, + { + "epoch": 12.138178913738018, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 15197 + }, + { + "epoch": 12.138977635782748, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 15198 + }, + { + "epoch": 12.139776357827476, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 15199 + }, + { + "epoch": 12.140575079872205, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 15200 + }, + { + "epoch": 12.141373801916933, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 15201 + }, + { + "epoch": 12.142172523961662, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 15202 + }, + { + "epoch": 12.14297124600639, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 15203 + }, + { + "epoch": 12.143769968051119, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 15204 + }, + { + "epoch": 12.144568690095847, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 15205 + }, + { + "epoch": 12.145367412140574, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 15206 + }, + { + "epoch": 12.146166134185304, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 15207 + }, + { + "epoch": 12.146964856230031, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 15208 + }, + { + "epoch": 12.14776357827476, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 15209 + }, + { + "epoch": 12.148562300319488, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 15210 + }, + { + "epoch": 12.149361022364218, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 15211 + }, + { + "epoch": 12.150159744408946, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 15212 + }, + { + "epoch": 12.150958466453675, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 15213 + }, + { + "epoch": 12.151757188498403, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 15214 + }, + { + "epoch": 12.15255591054313, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 15215 + }, + { + "epoch": 12.15335463258786, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 15216 + }, + { + "epoch": 12.154153354632587, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 15217 + }, + { + "epoch": 12.154952076677317, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 15218 + }, + { + "epoch": 12.155750798722044, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 15219 + }, + { + "epoch": 12.156549520766774, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 15220 + }, + { + "epoch": 12.157348242811501, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 15221 + }, + { + "epoch": 12.15814696485623, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 15222 + }, + { + "epoch": 12.158945686900958, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 15223 + }, + { + "epoch": 12.159744408945686, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 15224 + }, + { + "epoch": 12.160543130990416, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 15225 + }, + { + "epoch": 12.161341853035143, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 15226 + }, + { + "epoch": 12.162140575079873, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 15227 + }, + { + "epoch": 12.1629392971246, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 15228 + }, + { + "epoch": 12.16373801916933, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 15229 + }, + { + "epoch": 12.164536741214057, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 15230 + }, + { + "epoch": 12.165335463258787, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 15231 + }, + { + "epoch": 12.166134185303514, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 15232 + }, + { + "epoch": 12.166932907348242, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 15233 + }, + { + "epoch": 12.167731629392971, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 15234 + }, + { + "epoch": 12.168530351437699, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 15235 + }, + { + "epoch": 12.169329073482428, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 15236 + }, + { + "epoch": 12.170127795527156, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 15237 + }, + { + "epoch": 12.170926517571885, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 15238 + }, + { + "epoch": 12.171725239616613, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 15239 + }, + { + "epoch": 12.172523961661343, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 15240 + }, + { + "epoch": 12.17332268370607, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 15241 + }, + { + "epoch": 12.1741214057508, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 15242 + }, + { + "epoch": 12.174920127795527, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 15243 + }, + { + "epoch": 12.175718849840255, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 15244 + }, + { + "epoch": 12.176517571884984, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 15245 + }, + { + "epoch": 12.177316293929712, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 15246 + }, + { + "epoch": 12.178115015974441, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 15247 + }, + { + "epoch": 12.178913738019169, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 15248 + }, + { + "epoch": 12.179712460063898, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 15249 + }, + { + "epoch": 12.180511182108626, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 15250 + }, + { + "epoch": 12.181309904153355, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 15251 + }, + { + "epoch": 12.182108626198083, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 15252 + }, + { + "epoch": 12.18290734824281, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 15253 + }, + { + "epoch": 12.18370607028754, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 15254 + }, + { + "epoch": 12.184504792332268, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 15255 + }, + { + "epoch": 12.185303514376997, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 15256 + }, + { + "epoch": 12.186102236421725, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 15257 + }, + { + "epoch": 12.186900958466454, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 15258 + }, + { + "epoch": 12.187699680511182, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15259 + }, + { + "epoch": 12.188498402555911, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 15260 + }, + { + "epoch": 12.189297124600639, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 15261 + }, + { + "epoch": 12.190095846645367, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 15262 + }, + { + "epoch": 12.190894568690096, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 15263 + }, + { + "epoch": 12.191693290734824, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 15264 + }, + { + "epoch": 12.192492012779553, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 15265 + }, + { + "epoch": 12.19329073482428, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 15266 + }, + { + "epoch": 12.19408945686901, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 15267 + }, + { + "epoch": 12.194888178913738, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 15268 + }, + { + "epoch": 12.195686900958467, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 15269 + }, + { + "epoch": 12.196485623003195, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 15270 + }, + { + "epoch": 12.197284345047922, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 15271 + }, + { + "epoch": 12.198083067092652, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 15272 + }, + { + "epoch": 12.19888178913738, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 15273 + }, + { + "epoch": 12.199680511182109, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 15274 + }, + { + "epoch": 12.200479233226837, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 15275 + }, + { + "epoch": 12.201277955271566, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 15276 + }, + { + "epoch": 12.202076677316294, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 15277 + }, + { + "epoch": 12.202875399361023, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 15278 + }, + { + "epoch": 12.20367412140575, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 15279 + }, + { + "epoch": 12.204472843450478, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15280 + }, + { + "epoch": 12.205271565495208, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 15281 + }, + { + "epoch": 12.206070287539935, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 15282 + }, + { + "epoch": 12.206869009584665, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 15283 + }, + { + "epoch": 12.207667731629392, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 15284 + }, + { + "epoch": 12.208466453674122, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 15285 + }, + { + "epoch": 12.20926517571885, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 15286 + }, + { + "epoch": 12.210063897763579, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 15287 + }, + { + "epoch": 12.210862619808307, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 15288 + }, + { + "epoch": 12.211661341853036, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 15289 + }, + { + "epoch": 12.212460063897764, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 15290 + }, + { + "epoch": 12.213258785942491, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 15291 + }, + { + "epoch": 12.21405750798722, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 15292 + }, + { + "epoch": 12.214856230031948, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 15293 + }, + { + "epoch": 12.215654952076678, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 15294 + }, + { + "epoch": 12.216453674121405, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 15295 + }, + { + "epoch": 12.217252396166135, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 15296 + }, + { + "epoch": 12.218051118210862, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 15297 + }, + { + "epoch": 12.218849840255592, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 15298 + }, + { + "epoch": 12.21964856230032, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 15299 + }, + { + "epoch": 12.220447284345047, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 15300 + }, + { + "epoch": 12.221246006389777, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 15301 + }, + { + "epoch": 12.222044728434504, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 15302 + }, + { + "epoch": 12.222843450479234, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 15303 + }, + { + "epoch": 12.223642172523961, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 15304 + }, + { + "epoch": 12.22444089456869, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 15305 + }, + { + "epoch": 12.225239616613418, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 15306 + }, + { + "epoch": 12.226038338658148, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 15307 + }, + { + "epoch": 12.226837060702875, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 15308 + }, + { + "epoch": 12.227635782747603, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 15309 + }, + { + "epoch": 12.228434504792332, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 15310 + }, + { + "epoch": 12.22923322683706, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 15311 + }, + { + "epoch": 12.23003194888179, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 15312 + }, + { + "epoch": 12.230830670926517, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 15313 + }, + { + "epoch": 12.231629392971247, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 15314 + }, + { + "epoch": 12.232428115015974, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 15315 + }, + { + "epoch": 12.233226837060704, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 15316 + }, + { + "epoch": 12.234025559105431, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 15317 + }, + { + "epoch": 12.23482428115016, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 15318 + }, + { + "epoch": 12.235623003194888, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 15319 + }, + { + "epoch": 12.236421725239616, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 15320 + }, + { + "epoch": 12.237220447284345, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 15321 + }, + { + "epoch": 12.238019169329073, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 15322 + }, + { + "epoch": 12.238817891373802, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 15323 + }, + { + "epoch": 12.23961661341853, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 15324 + }, + { + "epoch": 12.24041533546326, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 15325 + }, + { + "epoch": 12.241214057507987, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15326 + }, + { + "epoch": 12.242012779552716, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 15327 + }, + { + "epoch": 12.242811501597444, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 15328 + }, + { + "epoch": 12.243610223642172, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 15329 + }, + { + "epoch": 12.244408945686901, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 15330 + }, + { + "epoch": 12.245207667731629, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15331 + }, + { + "epoch": 12.246006389776358, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 15332 + }, + { + "epoch": 12.246805111821086, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 15333 + }, + { + "epoch": 12.247603833865815, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 15334 + }, + { + "epoch": 12.248402555910543, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 15335 + }, + { + "epoch": 12.249201277955272, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 15336 + }, + { + "epoch": 12.25, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 15337 + }, + { + "epoch": 12.250798722044728, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 15338 + }, + { + "epoch": 12.251597444089457, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 15339 + }, + { + "epoch": 12.252396166134185, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 15340 + }, + { + "epoch": 12.253194888178914, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 15341 + }, + { + "epoch": 12.253993610223642, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1116, + "step": 15342 + }, + { + "epoch": 12.254792332268371, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 15343 + }, + { + "epoch": 12.255591054313099, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 15344 + }, + { + "epoch": 12.256389776357828, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 15345 + }, + { + "epoch": 12.257188498402556, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 15346 + }, + { + "epoch": 12.257987220447284, + "grad_norm": 0.09375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 15347 + }, + { + "epoch": 12.258785942492013, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 15348 + }, + { + "epoch": 12.25958466453674, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 15349 + }, + { + "epoch": 12.26038338658147, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 15350 + }, + { + "epoch": 12.261182108626198, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 15351 + }, + { + "epoch": 12.261980830670927, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 15352 + }, + { + "epoch": 12.262779552715655, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 15353 + }, + { + "epoch": 12.263578274760384, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 15354 + }, + { + "epoch": 12.264376996805112, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 15355 + }, + { + "epoch": 12.26517571884984, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 15356 + }, + { + "epoch": 12.265974440894569, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 15357 + }, + { + "epoch": 12.266773162939296, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 15358 + }, + { + "epoch": 12.267571884984026, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 15359 + }, + { + "epoch": 12.268370607028753, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 15360 + }, + { + "epoch": 12.269169329073483, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 15361 + }, + { + "epoch": 12.26996805111821, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 15362 + }, + { + "epoch": 12.27076677316294, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 15363 + }, + { + "epoch": 12.271565495207668, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 15364 + }, + { + "epoch": 12.272364217252397, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 15365 + }, + { + "epoch": 12.273162939297125, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 15366 + }, + { + "epoch": 12.273961661341852, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15367 + }, + { + "epoch": 12.274760383386582, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 15368 + }, + { + "epoch": 12.27555910543131, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 15369 + }, + { + "epoch": 12.276357827476039, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 15370 + }, + { + "epoch": 12.277156549520766, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 15371 + }, + { + "epoch": 12.277955271565496, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15372 + }, + { + "epoch": 12.278753993610223, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 15373 + }, + { + "epoch": 12.279552715654953, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 15374 + }, + { + "epoch": 12.28035143769968, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15375 + }, + { + "epoch": 12.281150159744408, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 15376 + }, + { + "epoch": 12.281948881789138, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 15377 + }, + { + "epoch": 12.282747603833865, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 15378 + }, + { + "epoch": 12.283546325878595, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 15379 + }, + { + "epoch": 12.284345047923322, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 15380 + }, + { + "epoch": 12.285143769968052, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 15381 + }, + { + "epoch": 12.28594249201278, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 15382 + }, + { + "epoch": 12.286741214057509, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 15383 + }, + { + "epoch": 12.287539936102236, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 15384 + }, + { + "epoch": 12.288338658146964, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 15385 + }, + { + "epoch": 12.289137380191693, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 15386 + }, + { + "epoch": 12.289936102236421, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 15387 + }, + { + "epoch": 12.29073482428115, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 15388 + }, + { + "epoch": 12.291533546325878, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 15389 + }, + { + "epoch": 12.292332268370608, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 15390 + }, + { + "epoch": 12.293130990415335, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 15391 + }, + { + "epoch": 12.293929712460065, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 15392 + }, + { + "epoch": 12.294728434504792, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 15393 + }, + { + "epoch": 12.295527156549522, + "grad_norm": 0.30859375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 15394 + }, + { + "epoch": 12.29632587859425, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 15395 + }, + { + "epoch": 12.297124600638977, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 15396 + }, + { + "epoch": 12.297923322683706, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15397 + }, + { + "epoch": 12.298722044728434, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 15398 + }, + { + "epoch": 12.299520766773163, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 15399 + }, + { + "epoch": 12.300319488817891, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 15400 + }, + { + "epoch": 12.30111821086262, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 15401 + }, + { + "epoch": 12.301916932907348, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 15402 + }, + { + "epoch": 12.302715654952078, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 15403 + }, + { + "epoch": 12.303514376996805, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 15404 + }, + { + "epoch": 12.304313099041533, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 15405 + }, + { + "epoch": 12.305111821086262, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 15406 + }, + { + "epoch": 12.30591054313099, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 15407 + }, + { + "epoch": 12.30670926517572, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 15408 + }, + { + "epoch": 12.307507987220447, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 15409 + }, + { + "epoch": 12.308306709265176, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 15410 + }, + { + "epoch": 12.309105431309904, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 15411 + }, + { + "epoch": 12.309904153354633, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 15412 + }, + { + "epoch": 12.310702875399361, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 15413 + }, + { + "epoch": 12.311501597444089, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 15414 + }, + { + "epoch": 12.312300319488818, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 15415 + }, + { + "epoch": 12.313099041533546, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 15416 + }, + { + "epoch": 12.313897763578275, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 15417 + }, + { + "epoch": 12.314696485623003, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 15418 + }, + { + "epoch": 12.315495207667732, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 15419 + }, + { + "epoch": 12.31629392971246, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 15420 + }, + { + "epoch": 12.31709265175719, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 15421 + }, + { + "epoch": 12.317891373801917, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 15422 + }, + { + "epoch": 12.318690095846645, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 15423 + }, + { + "epoch": 12.319488817891374, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 15424 + }, + { + "epoch": 12.320287539936102, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 15425 + }, + { + "epoch": 12.321086261980831, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 15426 + }, + { + "epoch": 12.321884984025559, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 15427 + }, + { + "epoch": 12.322683706070288, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 15428 + }, + { + "epoch": 12.323482428115016, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15429 + }, + { + "epoch": 12.324281150159745, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 15430 + }, + { + "epoch": 12.325079872204473, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15431 + }, + { + "epoch": 12.3258785942492, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 15432 + }, + { + "epoch": 12.32667731629393, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 15433 + }, + { + "epoch": 12.327476038338657, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 15434 + }, + { + "epoch": 12.328274760383387, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 15435 + }, + { + "epoch": 12.329073482428115, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 15436 + }, + { + "epoch": 12.329872204472844, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 15437 + }, + { + "epoch": 12.330670926517572, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 15438 + }, + { + "epoch": 12.331469648562301, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 15439 + }, + { + "epoch": 12.332268370607029, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 15440 + }, + { + "epoch": 12.333067092651758, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 15441 + }, + { + "epoch": 12.333865814696486, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 15442 + }, + { + "epoch": 12.334664536741213, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15443 + }, + { + "epoch": 12.335463258785943, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 15444 + }, + { + "epoch": 12.33626198083067, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 15445 + }, + { + "epoch": 12.3370607028754, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 15446 + }, + { + "epoch": 12.337859424920127, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 15447 + }, + { + "epoch": 12.338658146964857, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 15448 + }, + { + "epoch": 12.339456869009584, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 15449 + }, + { + "epoch": 12.340255591054314, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 15450 + }, + { + "epoch": 12.341054313099042, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 15451 + }, + { + "epoch": 12.34185303514377, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 15452 + }, + { + "epoch": 12.342651757188499, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 15453 + }, + { + "epoch": 12.343450479233226, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 15454 + }, + { + "epoch": 12.344249201277956, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 15455 + }, + { + "epoch": 12.345047923322683, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 15456 + }, + { + "epoch": 12.345846645367413, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 15457 + }, + { + "epoch": 12.34664536741214, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 15458 + }, + { + "epoch": 12.34744408945687, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 15459 + }, + { + "epoch": 12.348242811501597, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 15460 + }, + { + "epoch": 12.349041533546325, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 15461 + }, + { + "epoch": 12.349840255591054, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 15462 + }, + { + "epoch": 12.350638977635782, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 15463 + }, + { + "epoch": 12.351437699680512, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 15464 + }, + { + "epoch": 12.35223642172524, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 15465 + }, + { + "epoch": 12.353035143769969, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 15466 + }, + { + "epoch": 12.353833865814696, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 15467 + }, + { + "epoch": 12.354632587859426, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 15468 + }, + { + "epoch": 12.355431309904153, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 15469 + }, + { + "epoch": 12.356230031948883, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 15470 + }, + { + "epoch": 12.35702875399361, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 15471 + }, + { + "epoch": 12.357827476038338, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15472 + }, + { + "epoch": 12.358626198083067, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 15473 + }, + { + "epoch": 12.359424920127795, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 15474 + }, + { + "epoch": 12.360223642172524, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 15475 + }, + { + "epoch": 12.361022364217252, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 15476 + }, + { + "epoch": 12.361821086261982, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 15477 + }, + { + "epoch": 12.36261980830671, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 15478 + }, + { + "epoch": 12.363418530351439, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 15479 + }, + { + "epoch": 12.364217252396166, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 15480 + }, + { + "epoch": 12.365015974440894, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15481 + }, + { + "epoch": 12.365814696485623, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 15482 + }, + { + "epoch": 12.366613418530351, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 15483 + }, + { + "epoch": 12.36741214057508, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 15484 + }, + { + "epoch": 12.368210862619808, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 15485 + }, + { + "epoch": 12.369009584664537, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 15486 + }, + { + "epoch": 12.369808306709265, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 15487 + }, + { + "epoch": 12.370607028753994, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 15488 + }, + { + "epoch": 12.371405750798722, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 15489 + }, + { + "epoch": 12.37220447284345, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 15490 + }, + { + "epoch": 12.37300319488818, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 15491 + }, + { + "epoch": 12.373801916932907, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 15492 + }, + { + "epoch": 12.374600638977636, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 15493 + }, + { + "epoch": 12.375399361022364, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 15494 + }, + { + "epoch": 12.376198083067093, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 15495 + }, + { + "epoch": 12.37699680511182, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 15496 + }, + { + "epoch": 12.37779552715655, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 15497 + }, + { + "epoch": 12.378594249201278, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 15498 + }, + { + "epoch": 12.379392971246006, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 15499 + }, + { + "epoch": 12.380191693290735, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 15500 + }, + { + "epoch": 12.380990415335463, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1134, + "step": 15501 + }, + { + "epoch": 12.381789137380192, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 15502 + }, + { + "epoch": 12.38258785942492, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 15503 + }, + { + "epoch": 12.383386581469649, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 15504 + }, + { + "epoch": 12.384185303514377, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 15505 + }, + { + "epoch": 12.384984025559106, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 15506 + }, + { + "epoch": 12.385782747603834, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 15507 + }, + { + "epoch": 12.386581469648561, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 15508 + }, + { + "epoch": 12.38738019169329, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 15509 + }, + { + "epoch": 12.388178913738018, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 15510 + }, + { + "epoch": 12.388977635782748, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15511 + }, + { + "epoch": 12.389776357827476, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 15512 + }, + { + "epoch": 12.390575079872205, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 15513 + }, + { + "epoch": 12.391373801916933, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 15514 + }, + { + "epoch": 12.392172523961662, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 15515 + }, + { + "epoch": 12.39297124600639, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1127, + "step": 15516 + }, + { + "epoch": 12.393769968051117, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 15517 + }, + { + "epoch": 12.394568690095847, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 15518 + }, + { + "epoch": 12.395367412140574, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 15519 + }, + { + "epoch": 12.396166134185304, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 15520 + }, + { + "epoch": 12.396964856230031, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 15521 + }, + { + "epoch": 12.39776357827476, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 15522 + }, + { + "epoch": 12.398562300319488, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 15523 + }, + { + "epoch": 12.399361022364218, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 15524 + }, + { + "epoch": 12.400159744408946, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 15525 + }, + { + "epoch": 12.400958466453675, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 15526 + }, + { + "epoch": 12.401757188498403, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 15527 + }, + { + "epoch": 12.40255591054313, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 15528 + }, + { + "epoch": 12.40335463258786, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 15529 + }, + { + "epoch": 12.404153354632587, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 15530 + }, + { + "epoch": 12.404952076677317, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 15531 + }, + { + "epoch": 12.405750798722044, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 15532 + }, + { + "epoch": 12.406549520766774, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 15533 + }, + { + "epoch": 12.407348242811501, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 15534 + }, + { + "epoch": 12.40814696485623, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 15535 + }, + { + "epoch": 12.408945686900958, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 15536 + }, + { + "epoch": 12.409744408945686, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 15537 + }, + { + "epoch": 12.410543130990416, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 15538 + }, + { + "epoch": 12.411341853035143, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 15539 + }, + { + "epoch": 12.412140575079873, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 15540 + }, + { + "epoch": 12.4129392971246, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15541 + }, + { + "epoch": 12.41373801916933, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15542 + }, + { + "epoch": 12.414536741214057, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15543 + }, + { + "epoch": 12.415335463258787, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 15544 + }, + { + "epoch": 12.416134185303514, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 15545 + }, + { + "epoch": 12.416932907348242, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 15546 + }, + { + "epoch": 12.417731629392971, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 15547 + }, + { + "epoch": 12.418530351437699, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 15548 + }, + { + "epoch": 12.419329073482428, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 15549 + }, + { + "epoch": 12.420127795527156, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 15550 + }, + { + "epoch": 12.420926517571885, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 15551 + }, + { + "epoch": 12.421725239616613, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15552 + }, + { + "epoch": 12.422523961661343, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 15553 + }, + { + "epoch": 12.42332268370607, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 15554 + }, + { + "epoch": 12.4241214057508, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 15555 + }, + { + "epoch": 12.424920127795527, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 15556 + }, + { + "epoch": 12.425718849840255, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 15557 + }, + { + "epoch": 12.426517571884984, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 15558 + }, + { + "epoch": 12.427316293929712, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 15559 + }, + { + "epoch": 12.428115015974441, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 15560 + }, + { + "epoch": 12.428913738019169, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 15561 + }, + { + "epoch": 12.429712460063898, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 15562 + }, + { + "epoch": 12.430511182108626, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 15563 + }, + { + "epoch": 12.431309904153355, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 15564 + }, + { + "epoch": 12.432108626198083, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 15565 + }, + { + "epoch": 12.43290734824281, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 15566 + }, + { + "epoch": 12.43370607028754, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 15567 + }, + { + "epoch": 12.434504792332268, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 15568 + }, + { + "epoch": 12.435303514376997, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 15569 + }, + { + "epoch": 12.436102236421725, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 15570 + }, + { + "epoch": 12.436900958466454, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 15571 + }, + { + "epoch": 12.437699680511182, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 15572 + }, + { + "epoch": 12.438498402555911, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 15573 + }, + { + "epoch": 12.439297124600639, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 15574 + }, + { + "epoch": 12.440095846645367, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 15575 + }, + { + "epoch": 12.440894568690096, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 15576 + }, + { + "epoch": 12.441693290734824, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 15577 + }, + { + "epoch": 12.442492012779553, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 15578 + }, + { + "epoch": 12.44329073482428, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 15579 + }, + { + "epoch": 12.44408945686901, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 15580 + }, + { + "epoch": 12.444888178913738, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 15581 + }, + { + "epoch": 12.445686900958467, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15582 + }, + { + "epoch": 12.446485623003195, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 15583 + }, + { + "epoch": 12.447284345047922, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 15584 + }, + { + "epoch": 12.448083067092652, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 15585 + }, + { + "epoch": 12.44888178913738, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 15586 + }, + { + "epoch": 12.449680511182109, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15587 + }, + { + "epoch": 12.450479233226837, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 15588 + }, + { + "epoch": 12.451277955271566, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 15589 + }, + { + "epoch": 12.452076677316294, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 15590 + }, + { + "epoch": 12.452875399361023, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 15591 + }, + { + "epoch": 12.45367412140575, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 15592 + }, + { + "epoch": 12.454472843450478, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 15593 + }, + { + "epoch": 12.455271565495208, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 15594 + }, + { + "epoch": 12.456070287539935, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 15595 + }, + { + "epoch": 12.456869009584665, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 15596 + }, + { + "epoch": 12.457667731629392, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 15597 + }, + { + "epoch": 12.458466453674122, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 15598 + }, + { + "epoch": 12.45926517571885, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 15599 + }, + { + "epoch": 12.460063897763579, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 15600 + }, + { + "epoch": 12.460862619808307, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15601 + }, + { + "epoch": 12.461661341853036, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 15602 + }, + { + "epoch": 12.462460063897764, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 15603 + }, + { + "epoch": 12.463258785942491, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 15604 + }, + { + "epoch": 12.46405750798722, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 15605 + }, + { + "epoch": 12.464856230031948, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15606 + }, + { + "epoch": 12.465654952076678, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 15607 + }, + { + "epoch": 12.466453674121405, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 15608 + }, + { + "epoch": 12.467252396166135, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 15609 + }, + { + "epoch": 12.468051118210862, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 15610 + }, + { + "epoch": 12.468849840255592, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 15611 + }, + { + "epoch": 12.46964856230032, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15612 + }, + { + "epoch": 12.470447284345047, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 15613 + }, + { + "epoch": 12.471246006389777, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 15614 + }, + { + "epoch": 12.472044728434504, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 15615 + }, + { + "epoch": 12.472843450479234, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 15616 + }, + { + "epoch": 12.473642172523961, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 15617 + }, + { + "epoch": 12.47444089456869, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 15618 + }, + { + "epoch": 12.475239616613418, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 15619 + }, + { + "epoch": 12.476038338658148, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 15620 + }, + { + "epoch": 12.476837060702875, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 15621 + }, + { + "epoch": 12.477635782747603, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 15622 + }, + { + "epoch": 12.478434504792332, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 15623 + }, + { + "epoch": 12.47923322683706, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 15624 + }, + { + "epoch": 12.48003194888179, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 15625 + }, + { + "epoch": 12.480830670926517, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 15626 + }, + { + "epoch": 12.481629392971247, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 15627 + }, + { + "epoch": 12.482428115015974, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 15628 + }, + { + "epoch": 12.483226837060704, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 15629 + }, + { + "epoch": 12.484025559105431, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 15630 + }, + { + "epoch": 12.48482428115016, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 15631 + }, + { + "epoch": 12.485623003194888, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 15632 + }, + { + "epoch": 12.486421725239616, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 15633 + }, + { + "epoch": 12.487220447284345, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 15634 + }, + { + "epoch": 12.488019169329073, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 15635 + }, + { + "epoch": 12.488817891373802, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 15636 + }, + { + "epoch": 12.48961661341853, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 15637 + }, + { + "epoch": 12.49041533546326, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 15638 + }, + { + "epoch": 12.491214057507987, + "grad_norm": 0.318359375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 15639 + }, + { + "epoch": 12.492012779552716, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 15640 + }, + { + "epoch": 12.492811501597444, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 15641 + }, + { + "epoch": 12.493610223642172, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 15642 + }, + { + "epoch": 12.494408945686901, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 15643 + }, + { + "epoch": 12.495207667731629, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 15644 + }, + { + "epoch": 12.496006389776358, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 15645 + }, + { + "epoch": 12.496805111821086, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 15646 + }, + { + "epoch": 12.497603833865815, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15647 + }, + { + "epoch": 12.498402555910543, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 15648 + }, + { + "epoch": 12.499201277955272, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 15649 + }, + { + "epoch": 12.5, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 15650 + }, + { + "epoch": 12.500798722044728, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 15651 + }, + { + "epoch": 12.501597444089457, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 15652 + }, + { + "epoch": 12.502396166134185, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 15653 + }, + { + "epoch": 12.503194888178914, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 15654 + }, + { + "epoch": 12.503993610223642, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 15655 + }, + { + "epoch": 12.504792332268371, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 15656 + }, + { + "epoch": 12.505591054313099, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 15657 + }, + { + "epoch": 12.506389776357828, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 15658 + }, + { + "epoch": 12.507188498402556, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 15659 + }, + { + "epoch": 12.507987220447284, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 15660 + }, + { + "epoch": 12.508785942492013, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15661 + }, + { + "epoch": 12.50958466453674, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 15662 + }, + { + "epoch": 12.51038338658147, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 15663 + }, + { + "epoch": 12.511182108626198, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 15664 + }, + { + "epoch": 12.511980830670927, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 15665 + }, + { + "epoch": 12.512779552715655, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 15666 + }, + { + "epoch": 12.513578274760384, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15667 + }, + { + "epoch": 12.514376996805112, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 15668 + }, + { + "epoch": 12.51517571884984, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 15669 + }, + { + "epoch": 12.515974440894569, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 15670 + }, + { + "epoch": 12.516773162939296, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 15671 + }, + { + "epoch": 12.517571884984026, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 15672 + }, + { + "epoch": 12.518370607028753, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 15673 + }, + { + "epoch": 12.519169329073483, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 15674 + }, + { + "epoch": 12.51996805111821, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 15675 + }, + { + "epoch": 12.52076677316294, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 15676 + }, + { + "epoch": 12.521565495207668, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 15677 + }, + { + "epoch": 12.522364217252395, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 15678 + }, + { + "epoch": 12.523162939297125, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 15679 + }, + { + "epoch": 12.523961661341852, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 15680 + }, + { + "epoch": 12.524760383386582, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 15681 + }, + { + "epoch": 12.52555910543131, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15682 + }, + { + "epoch": 12.526357827476039, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 15683 + }, + { + "epoch": 12.527156549520766, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 15684 + }, + { + "epoch": 12.527955271565496, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 15685 + }, + { + "epoch": 12.528753993610223, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 15686 + }, + { + "epoch": 12.529552715654953, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 15687 + }, + { + "epoch": 12.53035143769968, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 15688 + }, + { + "epoch": 12.531150159744408, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 15689 + }, + { + "epoch": 12.531948881789138, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 15690 + }, + { + "epoch": 12.532747603833865, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 15691 + }, + { + "epoch": 12.533546325878595, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 15692 + }, + { + "epoch": 12.534345047923322, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 15693 + }, + { + "epoch": 12.535143769968052, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 15694 + }, + { + "epoch": 12.53594249201278, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 15695 + }, + { + "epoch": 12.536741214057509, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 15696 + }, + { + "epoch": 12.537539936102236, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 15697 + }, + { + "epoch": 12.538338658146966, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 15698 + }, + { + "epoch": 12.539137380191693, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 15699 + }, + { + "epoch": 12.539936102236421, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 15700 + }, + { + "epoch": 12.54073482428115, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 15701 + }, + { + "epoch": 12.541533546325878, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15702 + }, + { + "epoch": 12.542332268370608, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 15703 + }, + { + "epoch": 12.543130990415335, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 15704 + }, + { + "epoch": 12.543929712460065, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 15705 + }, + { + "epoch": 12.544728434504792, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 15706 + }, + { + "epoch": 12.545527156549522, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15707 + }, + { + "epoch": 12.54632587859425, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 15708 + }, + { + "epoch": 12.547124600638977, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 15709 + }, + { + "epoch": 12.547923322683706, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15710 + }, + { + "epoch": 12.548722044728434, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15711 + }, + { + "epoch": 12.549520766773163, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 15712 + }, + { + "epoch": 12.550319488817891, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 15713 + }, + { + "epoch": 12.55111821086262, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 15714 + }, + { + "epoch": 12.551916932907348, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 15715 + }, + { + "epoch": 12.552715654952078, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 15716 + }, + { + "epoch": 12.553514376996805, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 15717 + }, + { + "epoch": 12.554313099041533, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 15718 + }, + { + "epoch": 12.555111821086262, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 15719 + }, + { + "epoch": 12.55591054313099, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 15720 + }, + { + "epoch": 12.55670926517572, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 15721 + }, + { + "epoch": 12.557507987220447, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 15722 + }, + { + "epoch": 12.558306709265176, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 15723 + }, + { + "epoch": 12.559105431309904, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 15724 + }, + { + "epoch": 12.559904153354633, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 15725 + }, + { + "epoch": 12.560702875399361, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15726 + }, + { + "epoch": 12.561501597444089, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 15727 + }, + { + "epoch": 12.562300319488818, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 15728 + }, + { + "epoch": 12.563099041533546, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 15729 + }, + { + "epoch": 12.563897763578275, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 15730 + }, + { + "epoch": 12.564696485623003, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 15731 + }, + { + "epoch": 12.565495207667732, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 15732 + }, + { + "epoch": 12.56629392971246, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 15733 + }, + { + "epoch": 12.56709265175719, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 15734 + }, + { + "epoch": 12.567891373801917, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 15735 + }, + { + "epoch": 12.568690095846645, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 15736 + }, + { + "epoch": 12.569488817891374, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 15737 + }, + { + "epoch": 12.570287539936102, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 15738 + }, + { + "epoch": 12.571086261980831, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 15739 + }, + { + "epoch": 12.571884984025559, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 15740 + }, + { + "epoch": 12.572683706070288, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 15741 + }, + { + "epoch": 12.573482428115016, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 15742 + }, + { + "epoch": 12.574281150159745, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 15743 + }, + { + "epoch": 12.575079872204473, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 15744 + }, + { + "epoch": 12.5758785942492, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 15745 + }, + { + "epoch": 12.57667731629393, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 15746 + }, + { + "epoch": 12.577476038338657, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 15747 + }, + { + "epoch": 12.578274760383387, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 15748 + }, + { + "epoch": 12.579073482428115, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 15749 + }, + { + "epoch": 12.579872204472844, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 15750 + }, + { + "epoch": 12.580670926517572, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 15751 + }, + { + "epoch": 12.581469648562301, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 15752 + }, + { + "epoch": 12.582268370607029, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 15753 + }, + { + "epoch": 12.583067092651756, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 15754 + }, + { + "epoch": 12.583865814696486, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 15755 + }, + { + "epoch": 12.584664536741213, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 15756 + }, + { + "epoch": 12.585463258785943, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 15757 + }, + { + "epoch": 12.58626198083067, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 15758 + }, + { + "epoch": 12.5870607028754, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 15759 + }, + { + "epoch": 12.587859424920127, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 15760 + }, + { + "epoch": 12.588658146964857, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 15761 + }, + { + "epoch": 12.589456869009584, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 15762 + }, + { + "epoch": 12.590255591054314, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15763 + }, + { + "epoch": 12.591054313099042, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 15764 + }, + { + "epoch": 12.59185303514377, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 15765 + }, + { + "epoch": 12.592651757188499, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 15766 + }, + { + "epoch": 12.593450479233226, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 15767 + }, + { + "epoch": 12.594249201277956, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 15768 + }, + { + "epoch": 12.595047923322683, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 15769 + }, + { + "epoch": 12.595846645367413, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 15770 + }, + { + "epoch": 12.59664536741214, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 15771 + }, + { + "epoch": 12.59744408945687, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 15772 + }, + { + "epoch": 12.598242811501597, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 15773 + }, + { + "epoch": 12.599041533546325, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 15774 + }, + { + "epoch": 12.599840255591054, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 15775 + }, + { + "epoch": 12.600638977635782, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 15776 + }, + { + "epoch": 12.601437699680512, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 15777 + }, + { + "epoch": 12.60223642172524, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 15778 + }, + { + "epoch": 12.603035143769969, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 15779 + }, + { + "epoch": 12.603833865814696, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 15780 + }, + { + "epoch": 12.604632587859426, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 15781 + }, + { + "epoch": 12.605431309904153, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 15782 + }, + { + "epoch": 12.606230031948883, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 15783 + }, + { + "epoch": 12.60702875399361, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15784 + }, + { + "epoch": 12.607827476038338, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 15785 + }, + { + "epoch": 12.608626198083067, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 15786 + }, + { + "epoch": 12.609424920127795, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15787 + }, + { + "epoch": 12.610223642172524, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 15788 + }, + { + "epoch": 12.611022364217252, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15789 + }, + { + "epoch": 12.611821086261982, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 15790 + }, + { + "epoch": 12.61261980830671, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 15791 + }, + { + "epoch": 12.613418530351439, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 15792 + }, + { + "epoch": 12.614217252396166, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 15793 + }, + { + "epoch": 12.615015974440894, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 15794 + }, + { + "epoch": 12.615814696485623, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 15795 + }, + { + "epoch": 12.616613418530351, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 15796 + }, + { + "epoch": 12.61741214057508, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 15797 + }, + { + "epoch": 12.618210862619808, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 15798 + }, + { + "epoch": 12.619009584664537, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 15799 + }, + { + "epoch": 12.619808306709265, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 15800 + }, + { + "epoch": 12.620607028753994, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 15801 + }, + { + "epoch": 12.621405750798722, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 15802 + }, + { + "epoch": 12.62220447284345, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 15803 + }, + { + "epoch": 12.62300319488818, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 15804 + }, + { + "epoch": 12.623801916932907, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 15805 + }, + { + "epoch": 12.624600638977636, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15806 + }, + { + "epoch": 12.625399361022364, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 15807 + }, + { + "epoch": 12.626198083067093, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 15808 + }, + { + "epoch": 12.62699680511182, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 15809 + }, + { + "epoch": 12.62779552715655, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 15810 + }, + { + "epoch": 12.628594249201278, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 15811 + }, + { + "epoch": 12.629392971246006, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 15812 + }, + { + "epoch": 12.630191693290735, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 15813 + }, + { + "epoch": 12.630990415335463, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 15814 + }, + { + "epoch": 12.631789137380192, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 15815 + }, + { + "epoch": 12.63258785942492, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 15816 + }, + { + "epoch": 12.633386581469649, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15817 + }, + { + "epoch": 12.634185303514377, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 15818 + }, + { + "epoch": 12.634984025559106, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 15819 + }, + { + "epoch": 12.635782747603834, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 15820 + }, + { + "epoch": 12.636581469648561, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 15821 + }, + { + "epoch": 12.63738019169329, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 15822 + }, + { + "epoch": 12.638178913738018, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 15823 + }, + { + "epoch": 12.638977635782748, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 15824 + }, + { + "epoch": 12.639776357827476, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 15825 + }, + { + "epoch": 12.640575079872205, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 15826 + }, + { + "epoch": 12.641373801916933, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 15827 + }, + { + "epoch": 12.642172523961662, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 15828 + }, + { + "epoch": 12.64297124600639, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 15829 + }, + { + "epoch": 12.643769968051117, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 15830 + }, + { + "epoch": 12.644568690095847, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 15831 + }, + { + "epoch": 12.645367412140574, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 15832 + }, + { + "epoch": 12.646166134185304, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 15833 + }, + { + "epoch": 12.646964856230031, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 15834 + }, + { + "epoch": 12.64776357827476, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 15835 + }, + { + "epoch": 12.648562300319488, + "grad_norm": 0.09814453125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 15836 + }, + { + "epoch": 12.649361022364218, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 15837 + }, + { + "epoch": 12.650159744408946, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 15838 + }, + { + "epoch": 12.650958466453675, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 15839 + }, + { + "epoch": 12.651757188498403, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 15840 + }, + { + "epoch": 12.65255591054313, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 15841 + }, + { + "epoch": 12.65335463258786, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 15842 + }, + { + "epoch": 12.654153354632587, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 15843 + }, + { + "epoch": 12.654952076677317, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 15844 + }, + { + "epoch": 12.655750798722044, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 15845 + }, + { + "epoch": 12.656549520766774, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 15846 + }, + { + "epoch": 12.657348242811501, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 15847 + }, + { + "epoch": 12.65814696485623, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 15848 + }, + { + "epoch": 12.658945686900958, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 15849 + }, + { + "epoch": 12.659744408945686, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 15850 + }, + { + "epoch": 12.660543130990416, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 15851 + }, + { + "epoch": 12.661341853035143, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 15852 + }, + { + "epoch": 12.662140575079873, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 15853 + }, + { + "epoch": 12.6629392971246, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 15854 + }, + { + "epoch": 12.66373801916933, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 15855 + }, + { + "epoch": 12.664536741214057, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 15856 + }, + { + "epoch": 12.665335463258787, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 15857 + }, + { + "epoch": 12.666134185303514, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 15858 + }, + { + "epoch": 12.666932907348244, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 15859 + }, + { + "epoch": 12.667731629392971, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 15860 + }, + { + "epoch": 12.668530351437699, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 15861 + }, + { + "epoch": 12.669329073482428, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 15862 + }, + { + "epoch": 12.670127795527156, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 15863 + }, + { + "epoch": 12.670926517571885, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 15864 + }, + { + "epoch": 12.671725239616613, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 15865 + }, + { + "epoch": 12.672523961661343, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 15866 + }, + { + "epoch": 12.67332268370607, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 15867 + }, + { + "epoch": 12.6741214057508, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 15868 + }, + { + "epoch": 12.674920127795527, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15869 + }, + { + "epoch": 12.675718849840255, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 15870 + }, + { + "epoch": 12.676517571884984, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 15871 + }, + { + "epoch": 12.677316293929712, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 15872 + }, + { + "epoch": 12.678115015974441, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 15873 + }, + { + "epoch": 12.678913738019169, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 15874 + }, + { + "epoch": 12.679712460063898, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 15875 + }, + { + "epoch": 12.680511182108626, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 15876 + }, + { + "epoch": 12.681309904153355, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 15877 + }, + { + "epoch": 12.682108626198083, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 15878 + }, + { + "epoch": 12.68290734824281, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 15879 + }, + { + "epoch": 12.68370607028754, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 15880 + }, + { + "epoch": 12.684504792332268, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 15881 + }, + { + "epoch": 12.685303514376997, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 15882 + }, + { + "epoch": 12.686102236421725, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 15883 + }, + { + "epoch": 12.686900958466454, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 15884 + }, + { + "epoch": 12.687699680511182, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 15885 + }, + { + "epoch": 12.688498402555911, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 15886 + }, + { + "epoch": 12.689297124600639, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 15887 + }, + { + "epoch": 12.690095846645367, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 15888 + }, + { + "epoch": 12.690894568690096, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 15889 + }, + { + "epoch": 12.691693290734824, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 15890 + }, + { + "epoch": 12.692492012779553, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 15891 + }, + { + "epoch": 12.69329073482428, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 15892 + }, + { + "epoch": 12.69408945686901, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 15893 + }, + { + "epoch": 12.694888178913738, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 15894 + }, + { + "epoch": 12.695686900958467, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 15895 + }, + { + "epoch": 12.696485623003195, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 15896 + }, + { + "epoch": 12.697284345047922, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 15897 + }, + { + "epoch": 12.698083067092652, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 15898 + }, + { + "epoch": 12.69888178913738, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 15899 + }, + { + "epoch": 12.699680511182109, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 15900 + }, + { + "epoch": 12.700479233226837, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 15901 + }, + { + "epoch": 12.701277955271566, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 15902 + }, + { + "epoch": 12.702076677316294, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 15903 + }, + { + "epoch": 12.702875399361023, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 15904 + }, + { + "epoch": 12.70367412140575, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 15905 + }, + { + "epoch": 12.704472843450478, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 15906 + }, + { + "epoch": 12.705271565495208, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 15907 + }, + { + "epoch": 12.706070287539935, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 15908 + }, + { + "epoch": 12.706869009584665, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 15909 + }, + { + "epoch": 12.707667731629392, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 15910 + }, + { + "epoch": 12.708466453674122, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 15911 + }, + { + "epoch": 12.70926517571885, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 15912 + }, + { + "epoch": 12.710063897763579, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 15913 + }, + { + "epoch": 12.710862619808307, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 15914 + }, + { + "epoch": 12.711661341853034, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 15915 + }, + { + "epoch": 12.712460063897764, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 15916 + }, + { + "epoch": 12.713258785942491, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 15917 + }, + { + "epoch": 12.71405750798722, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 15918 + }, + { + "epoch": 12.714856230031948, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 15919 + }, + { + "epoch": 12.715654952076678, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 15920 + }, + { + "epoch": 12.716453674121405, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 15921 + }, + { + "epoch": 12.717252396166135, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 15922 + }, + { + "epoch": 12.718051118210862, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 15923 + }, + { + "epoch": 12.718849840255592, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 15924 + }, + { + "epoch": 12.71964856230032, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 15925 + }, + { + "epoch": 12.720447284345047, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 15926 + }, + { + "epoch": 12.721246006389777, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 15927 + }, + { + "epoch": 12.722044728434504, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 15928 + }, + { + "epoch": 12.722843450479234, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 15929 + }, + { + "epoch": 12.723642172523961, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 15930 + }, + { + "epoch": 12.72444089456869, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 15931 + }, + { + "epoch": 12.725239616613418, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 15932 + }, + { + "epoch": 12.726038338658148, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 15933 + }, + { + "epoch": 12.726837060702875, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 15934 + }, + { + "epoch": 12.727635782747605, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 15935 + }, + { + "epoch": 12.728434504792332, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 15936 + }, + { + "epoch": 12.72923322683706, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 15937 + }, + { + "epoch": 12.73003194888179, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 15938 + }, + { + "epoch": 12.730830670926517, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 15939 + }, + { + "epoch": 12.731629392971247, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 15940 + }, + { + "epoch": 12.732428115015974, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 15941 + }, + { + "epoch": 12.733226837060704, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 15942 + }, + { + "epoch": 12.734025559105431, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 15943 + }, + { + "epoch": 12.73482428115016, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 15944 + }, + { + "epoch": 12.735623003194888, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 15945 + }, + { + "epoch": 12.736421725239616, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 15946 + }, + { + "epoch": 12.737220447284345, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15947 + }, + { + "epoch": 12.738019169329073, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 15948 + }, + { + "epoch": 12.738817891373802, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 15949 + }, + { + "epoch": 12.73961661341853, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 15950 + }, + { + "epoch": 12.74041533546326, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 15951 + }, + { + "epoch": 12.741214057507987, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 15952 + }, + { + "epoch": 12.742012779552716, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15953 + }, + { + "epoch": 12.742811501597444, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 15954 + }, + { + "epoch": 12.743610223642172, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15955 + }, + { + "epoch": 12.744408945686901, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 15956 + }, + { + "epoch": 12.745207667731629, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 15957 + }, + { + "epoch": 12.746006389776358, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 15958 + }, + { + "epoch": 12.746805111821086, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 15959 + }, + { + "epoch": 12.747603833865815, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 15960 + }, + { + "epoch": 12.748402555910543, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 15961 + }, + { + "epoch": 12.749201277955272, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 15962 + }, + { + "epoch": 12.75, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15963 + }, + { + "epoch": 12.750798722044728, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 15964 + }, + { + "epoch": 12.751597444089457, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 15965 + }, + { + "epoch": 12.752396166134185, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 15966 + }, + { + "epoch": 12.753194888178914, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 15967 + }, + { + "epoch": 12.753993610223642, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 15968 + }, + { + "epoch": 12.754792332268371, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 15969 + }, + { + "epoch": 12.755591054313099, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 15970 + }, + { + "epoch": 12.756389776357828, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 15971 + }, + { + "epoch": 12.757188498402556, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 15972 + }, + { + "epoch": 12.757987220447284, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 15973 + }, + { + "epoch": 12.758785942492013, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 15974 + }, + { + "epoch": 12.75958466453674, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 15975 + }, + { + "epoch": 12.76038338658147, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 15976 + }, + { + "epoch": 12.761182108626198, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 15977 + }, + { + "epoch": 12.761980830670927, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 15978 + }, + { + "epoch": 12.762779552715655, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 15979 + }, + { + "epoch": 12.763578274760384, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 15980 + }, + { + "epoch": 12.764376996805112, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 15981 + }, + { + "epoch": 12.76517571884984, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 15982 + }, + { + "epoch": 12.765974440894569, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15983 + }, + { + "epoch": 12.766773162939296, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 15984 + }, + { + "epoch": 12.767571884984026, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 15985 + }, + { + "epoch": 12.768370607028753, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 15986 + }, + { + "epoch": 12.769169329073483, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 15987 + }, + { + "epoch": 12.76996805111821, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 15988 + }, + { + "epoch": 12.77076677316294, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 15989 + }, + { + "epoch": 12.771565495207668, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 15990 + }, + { + "epoch": 12.772364217252395, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 15991 + }, + { + "epoch": 12.773162939297125, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 15992 + }, + { + "epoch": 12.773961661341852, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 15993 + }, + { + "epoch": 12.774760383386582, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 15994 + }, + { + "epoch": 12.77555910543131, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 15995 + }, + { + "epoch": 12.776357827476039, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 15996 + }, + { + "epoch": 12.777156549520766, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 15997 + }, + { + "epoch": 12.777955271565496, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.112, + "step": 15998 + }, + { + "epoch": 12.778753993610223, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 15999 + }, + { + "epoch": 12.779552715654953, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 16000 + }, + { + "epoch": 12.78035143769968, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 16001 + }, + { + "epoch": 12.781150159744408, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 16002 + }, + { + "epoch": 12.781948881789138, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 16003 + }, + { + "epoch": 12.782747603833865, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 16004 + }, + { + "epoch": 12.783546325878595, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.109, + "step": 16005 + }, + { + "epoch": 12.784345047923322, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 16006 + }, + { + "epoch": 12.785143769968052, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 16007 + }, + { + "epoch": 12.78594249201278, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 16008 + }, + { + "epoch": 12.786741214057509, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 16009 + }, + { + "epoch": 12.787539936102236, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 16010 + }, + { + "epoch": 12.788338658146966, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 16011 + }, + { + "epoch": 12.789137380191693, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 16012 + }, + { + "epoch": 12.789936102236421, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 16013 + }, + { + "epoch": 12.79073482428115, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 16014 + }, + { + "epoch": 12.791533546325878, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 16015 + }, + { + "epoch": 12.792332268370608, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 16016 + }, + { + "epoch": 12.793130990415335, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 16017 + }, + { + "epoch": 12.793929712460065, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16018 + }, + { + "epoch": 12.794728434504792, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 16019 + }, + { + "epoch": 12.795527156549522, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1114, + "step": 16020 + }, + { + "epoch": 12.79632587859425, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 16021 + }, + { + "epoch": 12.797124600638977, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 16022 + }, + { + "epoch": 12.797923322683706, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 16023 + }, + { + "epoch": 12.798722044728434, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 16024 + }, + { + "epoch": 12.799520766773163, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 16025 + }, + { + "epoch": 12.800319488817891, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 16026 + }, + { + "epoch": 12.80111821086262, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 16027 + }, + { + "epoch": 12.801916932907348, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 16028 + }, + { + "epoch": 12.802715654952078, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 16029 + }, + { + "epoch": 12.803514376996805, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 16030 + }, + { + "epoch": 12.804313099041533, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 16031 + }, + { + "epoch": 12.805111821086262, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 16032 + }, + { + "epoch": 12.80591054313099, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 16033 + }, + { + "epoch": 12.80670926517572, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 16034 + }, + { + "epoch": 12.807507987220447, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 16035 + }, + { + "epoch": 12.808306709265176, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 16036 + }, + { + "epoch": 12.809105431309904, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 16037 + }, + { + "epoch": 12.809904153354633, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 16038 + }, + { + "epoch": 12.810702875399361, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 16039 + }, + { + "epoch": 12.811501597444089, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 16040 + }, + { + "epoch": 12.812300319488818, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16041 + }, + { + "epoch": 12.813099041533546, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1112, + "step": 16042 + }, + { + "epoch": 12.813897763578275, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 16043 + }, + { + "epoch": 12.814696485623003, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 16044 + }, + { + "epoch": 12.815495207667732, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 16045 + }, + { + "epoch": 12.81629392971246, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 16046 + }, + { + "epoch": 12.81709265175719, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 16047 + }, + { + "epoch": 12.817891373801917, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 16048 + }, + { + "epoch": 12.818690095846645, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 16049 + }, + { + "epoch": 12.819488817891374, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 16050 + }, + { + "epoch": 12.820287539936102, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 16051 + }, + { + "epoch": 12.821086261980831, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 16052 + }, + { + "epoch": 12.821884984025559, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16053 + }, + { + "epoch": 12.822683706070288, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.114, + "step": 16054 + }, + { + "epoch": 12.823482428115016, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16055 + }, + { + "epoch": 12.824281150159745, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 16056 + }, + { + "epoch": 12.825079872204473, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 16057 + }, + { + "epoch": 12.8258785942492, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 16058 + }, + { + "epoch": 12.82667731629393, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16059 + }, + { + "epoch": 12.827476038338657, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 16060 + }, + { + "epoch": 12.828274760383387, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 16061 + }, + { + "epoch": 12.829073482428115, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 16062 + }, + { + "epoch": 12.829872204472844, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 16063 + }, + { + "epoch": 12.830670926517572, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 16064 + }, + { + "epoch": 12.831469648562301, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 16065 + }, + { + "epoch": 12.832268370607029, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 16066 + }, + { + "epoch": 12.833067092651756, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 16067 + }, + { + "epoch": 12.833865814696486, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 16068 + }, + { + "epoch": 12.834664536741213, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16069 + }, + { + "epoch": 12.835463258785943, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 16070 + }, + { + "epoch": 12.83626198083067, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 16071 + }, + { + "epoch": 12.8370607028754, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 16072 + }, + { + "epoch": 12.837859424920127, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16073 + }, + { + "epoch": 12.838658146964857, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 16074 + }, + { + "epoch": 12.839456869009584, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 16075 + }, + { + "epoch": 12.840255591054314, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16076 + }, + { + "epoch": 12.841054313099042, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 16077 + }, + { + "epoch": 12.84185303514377, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 16078 + }, + { + "epoch": 12.842651757188499, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 16079 + }, + { + "epoch": 12.843450479233226, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 16080 + }, + { + "epoch": 12.844249201277956, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16081 + }, + { + "epoch": 12.845047923322683, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 16082 + }, + { + "epoch": 12.845846645367413, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 16083 + }, + { + "epoch": 12.84664536741214, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 16084 + }, + { + "epoch": 12.84744408945687, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1103, + "step": 16085 + }, + { + "epoch": 12.848242811501597, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 16086 + }, + { + "epoch": 12.849041533546325, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16087 + }, + { + "epoch": 12.849840255591054, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 16088 + }, + { + "epoch": 12.850638977635782, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 16089 + }, + { + "epoch": 12.851437699680512, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 16090 + }, + { + "epoch": 12.85223642172524, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 16091 + }, + { + "epoch": 12.853035143769969, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 16092 + }, + { + "epoch": 12.853833865814696, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 16093 + }, + { + "epoch": 12.854632587859426, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16094 + }, + { + "epoch": 12.855431309904153, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 16095 + }, + { + "epoch": 12.856230031948883, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 16096 + }, + { + "epoch": 12.85702875399361, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 16097 + }, + { + "epoch": 12.857827476038338, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16098 + }, + { + "epoch": 12.858626198083067, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 16099 + }, + { + "epoch": 12.859424920127795, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 16100 + }, + { + "epoch": 12.860223642172524, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 16101 + }, + { + "epoch": 12.861022364217252, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 16102 + }, + { + "epoch": 12.861821086261982, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 16103 + }, + { + "epoch": 12.86261980830671, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16104 + }, + { + "epoch": 12.863418530351439, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 16105 + }, + { + "epoch": 12.864217252396166, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 16106 + }, + { + "epoch": 12.865015974440894, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 16107 + }, + { + "epoch": 12.865814696485623, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 16108 + }, + { + "epoch": 12.866613418530351, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 16109 + }, + { + "epoch": 12.86741214057508, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 16110 + }, + { + "epoch": 12.868210862619808, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 16111 + }, + { + "epoch": 12.869009584664537, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 16112 + }, + { + "epoch": 12.869808306709265, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 16113 + }, + { + "epoch": 12.870607028753994, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 16114 + }, + { + "epoch": 12.871405750798722, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 16115 + }, + { + "epoch": 12.87220447284345, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 16116 + }, + { + "epoch": 12.87300319488818, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 16117 + }, + { + "epoch": 12.873801916932907, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 16118 + }, + { + "epoch": 12.874600638977636, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 16119 + }, + { + "epoch": 12.875399361022364, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 16120 + }, + { + "epoch": 12.876198083067093, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 16121 + }, + { + "epoch": 12.87699680511182, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 16122 + }, + { + "epoch": 12.87779552715655, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 16123 + }, + { + "epoch": 12.878594249201278, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 16124 + }, + { + "epoch": 12.879392971246006, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 16125 + }, + { + "epoch": 12.880191693290735, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 16126 + }, + { + "epoch": 12.880990415335463, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 16127 + }, + { + "epoch": 12.881789137380192, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 16128 + }, + { + "epoch": 12.88258785942492, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 16129 + }, + { + "epoch": 12.883386581469649, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 16130 + }, + { + "epoch": 12.884185303514377, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 16131 + }, + { + "epoch": 12.884984025559106, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16132 + }, + { + "epoch": 12.885782747603834, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 16133 + }, + { + "epoch": 12.886581469648561, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 16134 + }, + { + "epoch": 12.88738019169329, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 16135 + }, + { + "epoch": 12.888178913738018, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 16136 + }, + { + "epoch": 12.888977635782748, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 16137 + }, + { + "epoch": 12.889776357827476, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 16138 + }, + { + "epoch": 12.890575079872205, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 16139 + }, + { + "epoch": 12.891373801916933, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 16140 + }, + { + "epoch": 12.892172523961662, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 16141 + }, + { + "epoch": 12.89297124600639, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 16142 + }, + { + "epoch": 12.893769968051117, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 16143 + }, + { + "epoch": 12.894568690095847, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 16144 + }, + { + "epoch": 12.895367412140574, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 16145 + }, + { + "epoch": 12.896166134185304, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 16146 + }, + { + "epoch": 12.896964856230031, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 16147 + }, + { + "epoch": 12.89776357827476, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 16148 + }, + { + "epoch": 12.898562300319488, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 16149 + }, + { + "epoch": 12.899361022364218, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 16150 + }, + { + "epoch": 12.900159744408946, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 16151 + }, + { + "epoch": 12.900958466453675, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 16152 + }, + { + "epoch": 12.901757188498403, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 16153 + }, + { + "epoch": 12.90255591054313, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16154 + }, + { + "epoch": 12.90335463258786, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 16155 + }, + { + "epoch": 12.904153354632587, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 16156 + }, + { + "epoch": 12.904952076677317, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 16157 + }, + { + "epoch": 12.905750798722044, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 16158 + }, + { + "epoch": 12.906549520766774, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 16159 + }, + { + "epoch": 12.907348242811501, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16160 + }, + { + "epoch": 12.90814696485623, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 16161 + }, + { + "epoch": 12.908945686900958, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 16162 + }, + { + "epoch": 12.909744408945686, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 16163 + }, + { + "epoch": 12.910543130990416, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 16164 + }, + { + "epoch": 12.911341853035143, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 16165 + }, + { + "epoch": 12.912140575079873, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 16166 + }, + { + "epoch": 12.9129392971246, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 16167 + }, + { + "epoch": 12.91373801916933, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 16168 + }, + { + "epoch": 12.914536741214057, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 16169 + }, + { + "epoch": 12.915335463258787, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 16170 + }, + { + "epoch": 12.916134185303514, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 16171 + }, + { + "epoch": 12.916932907348244, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 16172 + }, + { + "epoch": 12.917731629392971, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 16173 + }, + { + "epoch": 12.918530351437699, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 16174 + }, + { + "epoch": 12.919329073482428, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 16175 + }, + { + "epoch": 12.920127795527156, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 16176 + }, + { + "epoch": 12.920926517571885, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 16177 + }, + { + "epoch": 12.921725239616613, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 16178 + }, + { + "epoch": 12.922523961661343, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 16179 + }, + { + "epoch": 12.92332268370607, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 16180 + }, + { + "epoch": 12.9241214057508, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 16181 + }, + { + "epoch": 12.924920127795527, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 16182 + }, + { + "epoch": 12.925718849840255, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 16183 + }, + { + "epoch": 12.926517571884984, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 16184 + }, + { + "epoch": 12.927316293929712, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 16185 + }, + { + "epoch": 12.928115015974441, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 16186 + }, + { + "epoch": 12.928913738019169, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 16187 + }, + { + "epoch": 12.929712460063898, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 16188 + }, + { + "epoch": 12.930511182108626, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 16189 + }, + { + "epoch": 12.931309904153355, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 16190 + }, + { + "epoch": 12.932108626198083, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 16191 + }, + { + "epoch": 12.93290734824281, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 16192 + }, + { + "epoch": 12.93370607028754, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 16193 + }, + { + "epoch": 12.934504792332268, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 16194 + }, + { + "epoch": 12.935303514376997, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 16195 + }, + { + "epoch": 12.936102236421725, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 16196 + }, + { + "epoch": 12.936900958466454, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 16197 + }, + { + "epoch": 12.937699680511182, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 16198 + }, + { + "epoch": 12.938498402555911, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 16199 + }, + { + "epoch": 12.939297124600639, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 16200 + }, + { + "epoch": 12.940095846645367, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 16201 + }, + { + "epoch": 12.940894568690096, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 16202 + }, + { + "epoch": 12.941693290734824, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 16203 + }, + { + "epoch": 12.942492012779553, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16204 + }, + { + "epoch": 12.94329073482428, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 16205 + }, + { + "epoch": 12.94408945686901, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 16206 + }, + { + "epoch": 12.944888178913738, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 16207 + }, + { + "epoch": 12.945686900958467, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 16208 + }, + { + "epoch": 12.946485623003195, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 16209 + }, + { + "epoch": 12.947284345047922, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 16210 + }, + { + "epoch": 12.948083067092652, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 16211 + }, + { + "epoch": 12.94888178913738, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 16212 + }, + { + "epoch": 12.949680511182109, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 16213 + }, + { + "epoch": 12.950479233226837, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16214 + }, + { + "epoch": 12.951277955271566, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 16215 + }, + { + "epoch": 12.952076677316294, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16216 + }, + { + "epoch": 12.952875399361023, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 16217 + }, + { + "epoch": 12.95367412140575, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 16218 + }, + { + "epoch": 12.954472843450478, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 16219 + }, + { + "epoch": 12.955271565495208, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 16220 + }, + { + "epoch": 12.956070287539935, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 16221 + }, + { + "epoch": 12.956869009584665, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 16222 + }, + { + "epoch": 12.957667731629392, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 16223 + }, + { + "epoch": 12.958466453674122, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 16224 + }, + { + "epoch": 12.95926517571885, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 16225 + }, + { + "epoch": 12.960063897763579, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16226 + }, + { + "epoch": 12.960862619808307, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 16227 + }, + { + "epoch": 12.961661341853034, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 16228 + }, + { + "epoch": 12.962460063897764, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 16229 + }, + { + "epoch": 12.963258785942491, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 16230 + }, + { + "epoch": 12.96405750798722, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 16231 + }, + { + "epoch": 12.964856230031948, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 16232 + }, + { + "epoch": 12.965654952076678, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 16233 + }, + { + "epoch": 12.966453674121405, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 16234 + }, + { + "epoch": 12.967252396166135, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 16235 + }, + { + "epoch": 12.968051118210862, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 16236 + }, + { + "epoch": 12.968849840255592, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 16237 + }, + { + "epoch": 12.96964856230032, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 16238 + }, + { + "epoch": 12.970447284345047, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 16239 + }, + { + "epoch": 12.971246006389777, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 16240 + }, + { + "epoch": 12.972044728434504, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 16241 + }, + { + "epoch": 12.972843450479234, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 16242 + }, + { + "epoch": 12.973642172523961, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 16243 + }, + { + "epoch": 12.97444089456869, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 16244 + }, + { + "epoch": 12.975239616613418, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 16245 + }, + { + "epoch": 12.976038338658148, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 16246 + }, + { + "epoch": 12.976837060702875, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 16247 + }, + { + "epoch": 12.977635782747605, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 16248 + }, + { + "epoch": 12.978434504792332, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 16249 + }, + { + "epoch": 12.97923322683706, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 16250 + }, + { + "epoch": 12.98003194888179, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 16251 + }, + { + "epoch": 12.980830670926517, + "grad_norm": 0.328125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 16252 + }, + { + "epoch": 12.981629392971247, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 16253 + }, + { + "epoch": 12.982428115015974, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 16254 + }, + { + "epoch": 12.983226837060704, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 16255 + }, + { + "epoch": 12.984025559105431, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 16256 + }, + { + "epoch": 12.98482428115016, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 16257 + }, + { + "epoch": 12.985623003194888, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 16258 + }, + { + "epoch": 12.986421725239616, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 16259 + }, + { + "epoch": 12.987220447284345, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 16260 + }, + { + "epoch": 12.988019169329073, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 16261 + }, + { + "epoch": 12.988817891373802, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 16262 + }, + { + "epoch": 12.98961661341853, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 16263 + }, + { + "epoch": 12.99041533546326, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 16264 + }, + { + "epoch": 12.991214057507987, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16265 + }, + { + "epoch": 12.992012779552716, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 16266 + }, + { + "epoch": 12.992811501597444, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1126, + "step": 16267 + }, + { + "epoch": 12.993610223642172, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 16268 + }, + { + "epoch": 12.994408945686901, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 16269 + }, + { + "epoch": 12.995207667731629, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 16270 + }, + { + "epoch": 12.996006389776358, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 16271 + }, + { + "epoch": 12.996805111821086, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 16272 + }, + { + "epoch": 12.997603833865815, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 16273 + }, + { + "epoch": 12.998402555910543, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16274 + }, + { + "epoch": 12.999201277955272, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 16275 + }, + { + "epoch": 13.0, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 16276 + }, + { + "epoch": 13.000798722044728, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1101, + "step": 16277 + }, + { + "epoch": 13.001597444089457, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 16278 + }, + { + "epoch": 13.002396166134185, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 16279 + }, + { + "epoch": 13.003194888178914, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16280 + }, + { + "epoch": 13.003993610223642, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 16281 + }, + { + "epoch": 13.004792332268371, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 16282 + }, + { + "epoch": 13.005591054313099, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 16283 + }, + { + "epoch": 13.006389776357828, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 16284 + }, + { + "epoch": 13.007188498402556, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16285 + }, + { + "epoch": 13.007987220447284, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 16286 + }, + { + "epoch": 13.008785942492013, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 16287 + }, + { + "epoch": 13.00958466453674, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 16288 + }, + { + "epoch": 13.01038338658147, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 16289 + }, + { + "epoch": 13.011182108626198, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 16290 + }, + { + "epoch": 13.011980830670927, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 16291 + }, + { + "epoch": 13.012779552715655, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16292 + }, + { + "epoch": 13.013578274760384, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 16293 + }, + { + "epoch": 13.014376996805112, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16294 + }, + { + "epoch": 13.01517571884984, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 16295 + }, + { + "epoch": 13.015974440894569, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 16296 + }, + { + "epoch": 13.016773162939296, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16297 + }, + { + "epoch": 13.017571884984026, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 16298 + }, + { + "epoch": 13.018370607028753, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 16299 + }, + { + "epoch": 13.019169329073483, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 16300 + }, + { + "epoch": 13.01996805111821, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 16301 + }, + { + "epoch": 13.02076677316294, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 16302 + }, + { + "epoch": 13.021565495207668, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 16303 + }, + { + "epoch": 13.022364217252397, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 16304 + }, + { + "epoch": 13.023162939297125, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16305 + }, + { + "epoch": 13.023961661341852, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16306 + }, + { + "epoch": 13.024760383386582, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 16307 + }, + { + "epoch": 13.02555910543131, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 16308 + }, + { + "epoch": 13.026357827476039, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 16309 + }, + { + "epoch": 13.027156549520766, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 16310 + }, + { + "epoch": 13.027955271565496, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 16311 + }, + { + "epoch": 13.028753993610223, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 16312 + }, + { + "epoch": 13.029552715654953, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 16313 + }, + { + "epoch": 13.03035143769968, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 16314 + }, + { + "epoch": 13.031150159744408, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 16315 + }, + { + "epoch": 13.031948881789138, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 16316 + }, + { + "epoch": 13.032747603833865, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 16317 + }, + { + "epoch": 13.033546325878595, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 16318 + }, + { + "epoch": 13.034345047923322, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 16319 + }, + { + "epoch": 13.035143769968052, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 16320 + }, + { + "epoch": 13.03594249201278, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 16321 + }, + { + "epoch": 13.036741214057509, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16322 + }, + { + "epoch": 13.037539936102236, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 16323 + }, + { + "epoch": 13.038338658146964, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 16324 + }, + { + "epoch": 13.039137380191693, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 16325 + }, + { + "epoch": 13.039936102236421, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 16326 + }, + { + "epoch": 13.04073482428115, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 16327 + }, + { + "epoch": 13.041533546325878, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 16328 + }, + { + "epoch": 13.042332268370608, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 16329 + }, + { + "epoch": 13.043130990415335, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 16330 + }, + { + "epoch": 13.043929712460065, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 16331 + }, + { + "epoch": 13.044728434504792, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 16332 + }, + { + "epoch": 13.04552715654952, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 16333 + }, + { + "epoch": 13.04632587859425, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 16334 + }, + { + "epoch": 13.047124600638977, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 16335 + }, + { + "epoch": 13.047923322683706, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16336 + }, + { + "epoch": 13.048722044728434, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 16337 + }, + { + "epoch": 13.049520766773163, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 16338 + }, + { + "epoch": 13.050319488817891, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 16339 + }, + { + "epoch": 13.05111821086262, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 16340 + }, + { + "epoch": 13.051916932907348, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 16341 + }, + { + "epoch": 13.052715654952078, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 16342 + }, + { + "epoch": 13.053514376996805, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16343 + }, + { + "epoch": 13.054313099041533, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 16344 + }, + { + "epoch": 13.055111821086262, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 16345 + }, + { + "epoch": 13.05591054313099, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 16346 + }, + { + "epoch": 13.05670926517572, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 16347 + }, + { + "epoch": 13.057507987220447, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 16348 + }, + { + "epoch": 13.058306709265176, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 16349 + }, + { + "epoch": 13.059105431309904, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 16350 + }, + { + "epoch": 13.059904153354633, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 16351 + }, + { + "epoch": 13.060702875399361, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 16352 + }, + { + "epoch": 13.061501597444089, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 16353 + }, + { + "epoch": 13.062300319488818, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 16354 + }, + { + "epoch": 13.063099041533546, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 16355 + }, + { + "epoch": 13.063897763578275, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 16356 + }, + { + "epoch": 13.064696485623003, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 16357 + }, + { + "epoch": 13.065495207667732, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 16358 + }, + { + "epoch": 13.06629392971246, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 16359 + }, + { + "epoch": 13.06709265175719, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 16360 + }, + { + "epoch": 13.067891373801917, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 16361 + }, + { + "epoch": 13.068690095846645, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 16362 + }, + { + "epoch": 13.069488817891374, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 16363 + }, + { + "epoch": 13.070287539936102, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 16364 + }, + { + "epoch": 13.071086261980831, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 16365 + }, + { + "epoch": 13.071884984025559, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 16366 + }, + { + "epoch": 13.072683706070288, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 16367 + }, + { + "epoch": 13.073482428115016, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 16368 + }, + { + "epoch": 13.074281150159745, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 16369 + }, + { + "epoch": 13.075079872204473, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 16370 + }, + { + "epoch": 13.0758785942492, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 16371 + }, + { + "epoch": 13.07667731629393, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 16372 + }, + { + "epoch": 13.077476038338657, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 16373 + }, + { + "epoch": 13.078274760383387, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 16374 + }, + { + "epoch": 13.079073482428115, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 16375 + }, + { + "epoch": 13.079872204472844, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 16376 + }, + { + "epoch": 13.080670926517572, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 16377 + }, + { + "epoch": 13.081469648562301, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 16378 + }, + { + "epoch": 13.082268370607029, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 16379 + }, + { + "epoch": 13.083067092651758, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 16380 + }, + { + "epoch": 13.083865814696486, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 16381 + }, + { + "epoch": 13.084664536741213, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 16382 + }, + { + "epoch": 13.085463258785943, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 16383 + }, + { + "epoch": 13.08626198083067, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 16384 + }, + { + "epoch": 13.0870607028754, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 16385 + }, + { + "epoch": 13.087859424920127, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 16386 + }, + { + "epoch": 13.088658146964857, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 16387 + }, + { + "epoch": 13.089456869009584, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 16388 + }, + { + "epoch": 13.090255591054314, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 16389 + }, + { + "epoch": 13.091054313099042, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16390 + }, + { + "epoch": 13.09185303514377, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 16391 + }, + { + "epoch": 13.092651757188499, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 16392 + }, + { + "epoch": 13.093450479233226, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 16393 + }, + { + "epoch": 13.094249201277956, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 16394 + }, + { + "epoch": 13.095047923322683, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 16395 + }, + { + "epoch": 13.095846645367413, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 16396 + }, + { + "epoch": 13.09664536741214, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 16397 + }, + { + "epoch": 13.09744408945687, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16398 + }, + { + "epoch": 13.098242811501597, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 16399 + }, + { + "epoch": 13.099041533546325, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 16400 + }, + { + "epoch": 13.099840255591054, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 16401 + }, + { + "epoch": 13.100638977635782, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 16402 + }, + { + "epoch": 13.101437699680512, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 16403 + }, + { + "epoch": 13.10223642172524, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 16404 + }, + { + "epoch": 13.103035143769969, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 16405 + }, + { + "epoch": 13.103833865814696, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 16406 + }, + { + "epoch": 13.104632587859426, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 16407 + }, + { + "epoch": 13.105431309904153, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 16408 + }, + { + "epoch": 13.106230031948881, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 16409 + }, + { + "epoch": 13.10702875399361, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 16410 + }, + { + "epoch": 13.107827476038338, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 16411 + }, + { + "epoch": 13.108626198083067, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 16412 + }, + { + "epoch": 13.109424920127795, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 16413 + }, + { + "epoch": 13.110223642172524, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 16414 + }, + { + "epoch": 13.111022364217252, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 16415 + }, + { + "epoch": 13.111821086261982, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1128, + "step": 16416 + }, + { + "epoch": 13.11261980830671, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 16417 + }, + { + "epoch": 13.113418530351439, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 16418 + }, + { + "epoch": 13.114217252396166, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 16419 + }, + { + "epoch": 13.115015974440894, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16420 + }, + { + "epoch": 13.115814696485623, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 16421 + }, + { + "epoch": 13.116613418530351, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 16422 + }, + { + "epoch": 13.11741214057508, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 16423 + }, + { + "epoch": 13.118210862619808, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 16424 + }, + { + "epoch": 13.119009584664537, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 16425 + }, + { + "epoch": 13.119808306709265, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 16426 + }, + { + "epoch": 13.120607028753994, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 16427 + }, + { + "epoch": 13.121405750798722, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 16428 + }, + { + "epoch": 13.12220447284345, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16429 + }, + { + "epoch": 13.12300319488818, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 16430 + }, + { + "epoch": 13.123801916932907, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 16431 + }, + { + "epoch": 13.124600638977636, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 16432 + }, + { + "epoch": 13.125399361022364, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 16433 + }, + { + "epoch": 13.126198083067093, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 16434 + }, + { + "epoch": 13.12699680511182, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 16435 + }, + { + "epoch": 13.12779552715655, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 16436 + }, + { + "epoch": 13.128594249201278, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 16437 + }, + { + "epoch": 13.129392971246006, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16438 + }, + { + "epoch": 13.130191693290735, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 16439 + }, + { + "epoch": 13.130990415335463, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 16440 + }, + { + "epoch": 13.131789137380192, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 16441 + }, + { + "epoch": 13.13258785942492, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 16442 + }, + { + "epoch": 13.133386581469649, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 16443 + }, + { + "epoch": 13.134185303514377, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 16444 + }, + { + "epoch": 13.134984025559106, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 16445 + }, + { + "epoch": 13.135782747603834, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 16446 + }, + { + "epoch": 13.136581469648561, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 16447 + }, + { + "epoch": 13.13738019169329, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 16448 + }, + { + "epoch": 13.138178913738018, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 16449 + }, + { + "epoch": 13.138977635782748, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 16450 + }, + { + "epoch": 13.139776357827476, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 16451 + }, + { + "epoch": 13.140575079872205, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 16452 + }, + { + "epoch": 13.141373801916933, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 16453 + }, + { + "epoch": 13.142172523961662, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1094, + "step": 16454 + }, + { + "epoch": 13.14297124600639, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 16455 + }, + { + "epoch": 13.143769968051119, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 16456 + }, + { + "epoch": 13.144568690095847, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 16457 + }, + { + "epoch": 13.145367412140574, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 16458 + }, + { + "epoch": 13.146166134185304, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 16459 + }, + { + "epoch": 13.146964856230031, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 16460 + }, + { + "epoch": 13.14776357827476, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 16461 + }, + { + "epoch": 13.148562300319488, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 16462 + }, + { + "epoch": 13.149361022364218, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 16463 + }, + { + "epoch": 13.150159744408946, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 16464 + }, + { + "epoch": 13.150958466453675, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 16465 + }, + { + "epoch": 13.151757188498403, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 16466 + }, + { + "epoch": 13.15255591054313, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 16467 + }, + { + "epoch": 13.15335463258786, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 16468 + }, + { + "epoch": 13.154153354632587, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 16469 + }, + { + "epoch": 13.154952076677317, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 16470 + }, + { + "epoch": 13.155750798722044, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 16471 + }, + { + "epoch": 13.156549520766774, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 16472 + }, + { + "epoch": 13.157348242811501, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 16473 + }, + { + "epoch": 13.15814696485623, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 16474 + }, + { + "epoch": 13.158945686900958, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 16475 + }, + { + "epoch": 13.159744408945686, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 16476 + }, + { + "epoch": 13.160543130990416, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 16477 + }, + { + "epoch": 13.161341853035143, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 16478 + }, + { + "epoch": 13.162140575079873, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 16479 + }, + { + "epoch": 13.1629392971246, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 16480 + }, + { + "epoch": 13.16373801916933, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 16481 + }, + { + "epoch": 13.164536741214057, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 16482 + }, + { + "epoch": 13.165335463258787, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 16483 + }, + { + "epoch": 13.166134185303514, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.113, + "step": 16484 + }, + { + "epoch": 13.166932907348242, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 16485 + }, + { + "epoch": 13.167731629392971, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 16486 + }, + { + "epoch": 13.168530351437699, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 16487 + }, + { + "epoch": 13.169329073482428, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 16488 + }, + { + "epoch": 13.170127795527156, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 16489 + }, + { + "epoch": 13.170926517571885, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 16490 + }, + { + "epoch": 13.171725239616613, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 16491 + }, + { + "epoch": 13.172523961661343, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 16492 + }, + { + "epoch": 13.17332268370607, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 16493 + }, + { + "epoch": 13.1741214057508, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 16494 + }, + { + "epoch": 13.174920127795527, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 16495 + }, + { + "epoch": 13.175718849840255, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 16496 + }, + { + "epoch": 13.176517571884984, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 16497 + }, + { + "epoch": 13.177316293929712, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16498 + }, + { + "epoch": 13.178115015974441, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 16499 + }, + { + "epoch": 13.178913738019169, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 16500 + }, + { + "epoch": 13.179712460063898, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 16501 + }, + { + "epoch": 13.180511182108626, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 16502 + }, + { + "epoch": 13.181309904153355, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 16503 + }, + { + "epoch": 13.182108626198083, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 16504 + }, + { + "epoch": 13.18290734824281, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 16505 + }, + { + "epoch": 13.18370607028754, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 16506 + }, + { + "epoch": 13.184504792332268, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 16507 + }, + { + "epoch": 13.185303514376997, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 16508 + }, + { + "epoch": 13.186102236421725, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 16509 + }, + { + "epoch": 13.186900958466454, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 16510 + }, + { + "epoch": 13.187699680511182, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 16511 + }, + { + "epoch": 13.188498402555911, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 16512 + }, + { + "epoch": 13.189297124600639, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 16513 + }, + { + "epoch": 13.190095846645367, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 16514 + }, + { + "epoch": 13.190894568690096, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 16515 + }, + { + "epoch": 13.191693290734824, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 16516 + }, + { + "epoch": 13.192492012779553, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 16517 + }, + { + "epoch": 13.19329073482428, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 16518 + }, + { + "epoch": 13.19408945686901, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 16519 + }, + { + "epoch": 13.194888178913738, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 16520 + }, + { + "epoch": 13.195686900958467, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 16521 + }, + { + "epoch": 13.196485623003195, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 16522 + }, + { + "epoch": 13.197284345047922, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 16523 + }, + { + "epoch": 13.198083067092652, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 16524 + }, + { + "epoch": 13.19888178913738, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 16525 + }, + { + "epoch": 13.199680511182109, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 16526 + }, + { + "epoch": 13.200479233226837, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16527 + }, + { + "epoch": 13.201277955271566, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 16528 + }, + { + "epoch": 13.202076677316294, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 16529 + }, + { + "epoch": 13.202875399361023, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 16530 + }, + { + "epoch": 13.20367412140575, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 16531 + }, + { + "epoch": 13.204472843450478, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 16532 + }, + { + "epoch": 13.205271565495208, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 16533 + }, + { + "epoch": 13.206070287539935, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 16534 + }, + { + "epoch": 13.206869009584665, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 16535 + }, + { + "epoch": 13.207667731629392, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 16536 + }, + { + "epoch": 13.208466453674122, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.111, + "step": 16537 + }, + { + "epoch": 13.20926517571885, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 16538 + }, + { + "epoch": 13.210063897763579, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 16539 + }, + { + "epoch": 13.210862619808307, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 16540 + }, + { + "epoch": 13.211661341853036, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 16541 + }, + { + "epoch": 13.212460063897764, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 16542 + }, + { + "epoch": 13.213258785942491, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 16543 + }, + { + "epoch": 13.21405750798722, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 16544 + }, + { + "epoch": 13.214856230031948, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 16545 + }, + { + "epoch": 13.215654952076678, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 16546 + }, + { + "epoch": 13.216453674121405, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 16547 + }, + { + "epoch": 13.217252396166135, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1107, + "step": 16548 + }, + { + "epoch": 13.218051118210862, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 16549 + }, + { + "epoch": 13.218849840255592, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 16550 + }, + { + "epoch": 13.21964856230032, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 16551 + }, + { + "epoch": 13.220447284345047, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 16552 + }, + { + "epoch": 13.221246006389777, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 16553 + }, + { + "epoch": 13.222044728434504, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16554 + }, + { + "epoch": 13.222843450479234, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 16555 + }, + { + "epoch": 13.223642172523961, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 16556 + }, + { + "epoch": 13.22444089456869, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 16557 + }, + { + "epoch": 13.225239616613418, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 16558 + }, + { + "epoch": 13.226038338658148, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 16559 + }, + { + "epoch": 13.226837060702875, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 16560 + }, + { + "epoch": 13.227635782747603, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 16561 + }, + { + "epoch": 13.228434504792332, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 16562 + }, + { + "epoch": 13.22923322683706, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 16563 + }, + { + "epoch": 13.23003194888179, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 16564 + }, + { + "epoch": 13.230830670926517, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 16565 + }, + { + "epoch": 13.231629392971247, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 16566 + }, + { + "epoch": 13.232428115015974, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 16567 + }, + { + "epoch": 13.233226837060704, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 16568 + }, + { + "epoch": 13.234025559105431, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 16569 + }, + { + "epoch": 13.23482428115016, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16570 + }, + { + "epoch": 13.235623003194888, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16571 + }, + { + "epoch": 13.236421725239616, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 16572 + }, + { + "epoch": 13.237220447284345, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 16573 + }, + { + "epoch": 13.238019169329073, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 16574 + }, + { + "epoch": 13.238817891373802, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 16575 + }, + { + "epoch": 13.23961661341853, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 16576 + }, + { + "epoch": 13.24041533546326, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 16577 + }, + { + "epoch": 13.241214057507987, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 16578 + }, + { + "epoch": 13.242012779552716, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 16579 + }, + { + "epoch": 13.242811501597444, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 16580 + }, + { + "epoch": 13.243610223642172, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16581 + }, + { + "epoch": 13.244408945686901, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 16582 + }, + { + "epoch": 13.245207667731629, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 16583 + }, + { + "epoch": 13.246006389776358, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 16584 + }, + { + "epoch": 13.246805111821086, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 16585 + }, + { + "epoch": 13.247603833865815, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 16586 + }, + { + "epoch": 13.248402555910543, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 16587 + }, + { + "epoch": 13.249201277955272, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 16588 + }, + { + "epoch": 13.25, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 16589 + }, + { + "epoch": 13.250798722044728, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 16590 + }, + { + "epoch": 13.251597444089457, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 16591 + }, + { + "epoch": 13.252396166134185, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 16592 + }, + { + "epoch": 13.253194888178914, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 16593 + }, + { + "epoch": 13.253993610223642, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 16594 + }, + { + "epoch": 13.254792332268371, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 16595 + }, + { + "epoch": 13.255591054313099, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 16596 + }, + { + "epoch": 13.256389776357828, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 16597 + }, + { + "epoch": 13.257188498402556, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 16598 + }, + { + "epoch": 13.257987220447284, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 16599 + }, + { + "epoch": 13.258785942492013, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 16600 + }, + { + "epoch": 13.25958466453674, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 16601 + }, + { + "epoch": 13.26038338658147, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 16602 + }, + { + "epoch": 13.261182108626198, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 16603 + }, + { + "epoch": 13.261980830670927, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 16604 + }, + { + "epoch": 13.262779552715655, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 16605 + }, + { + "epoch": 13.263578274760384, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 16606 + }, + { + "epoch": 13.264376996805112, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16607 + }, + { + "epoch": 13.26517571884984, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 16608 + }, + { + "epoch": 13.265974440894569, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 16609 + }, + { + "epoch": 13.266773162939296, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 16610 + }, + { + "epoch": 13.267571884984026, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 16611 + }, + { + "epoch": 13.268370607028753, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 16612 + }, + { + "epoch": 13.269169329073483, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 16613 + }, + { + "epoch": 13.26996805111821, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 16614 + }, + { + "epoch": 13.27076677316294, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16615 + }, + { + "epoch": 13.271565495207668, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 16616 + }, + { + "epoch": 13.272364217252397, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 16617 + }, + { + "epoch": 13.273162939297125, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 16618 + }, + { + "epoch": 13.273961661341852, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 16619 + }, + { + "epoch": 13.274760383386582, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 16620 + }, + { + "epoch": 13.27555910543131, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 16621 + }, + { + "epoch": 13.276357827476039, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 16622 + }, + { + "epoch": 13.277156549520766, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 16623 + }, + { + "epoch": 13.277955271565496, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 16624 + }, + { + "epoch": 13.278753993610223, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 16625 + }, + { + "epoch": 13.279552715654953, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 16626 + }, + { + "epoch": 13.28035143769968, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 16627 + }, + { + "epoch": 13.281150159744408, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 16628 + }, + { + "epoch": 13.281948881789138, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 16629 + }, + { + "epoch": 13.282747603833865, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 16630 + }, + { + "epoch": 13.283546325878595, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 16631 + }, + { + "epoch": 13.284345047923322, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 16632 + }, + { + "epoch": 13.285143769968052, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 16633 + }, + { + "epoch": 13.28594249201278, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 16634 + }, + { + "epoch": 13.286741214057509, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 16635 + }, + { + "epoch": 13.287539936102236, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 16636 + }, + { + "epoch": 13.288338658146964, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 16637 + }, + { + "epoch": 13.289137380191693, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 16638 + }, + { + "epoch": 13.289936102236421, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 16639 + }, + { + "epoch": 13.29073482428115, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 16640 + }, + { + "epoch": 13.291533546325878, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 16641 + }, + { + "epoch": 13.292332268370608, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 16642 + }, + { + "epoch": 13.293130990415335, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 16643 + }, + { + "epoch": 13.293929712460065, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 16644 + }, + { + "epoch": 13.294728434504792, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 16645 + }, + { + "epoch": 13.295527156549522, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 16646 + }, + { + "epoch": 13.29632587859425, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 16647 + }, + { + "epoch": 13.297124600638977, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 16648 + }, + { + "epoch": 13.297923322683706, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 16649 + }, + { + "epoch": 13.298722044728434, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 16650 + }, + { + "epoch": 13.299520766773163, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 16651 + }, + { + "epoch": 13.300319488817891, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 16652 + }, + { + "epoch": 13.30111821086262, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 16653 + }, + { + "epoch": 13.301916932907348, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 16654 + }, + { + "epoch": 13.302715654952078, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 16655 + }, + { + "epoch": 13.303514376996805, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 16656 + }, + { + "epoch": 13.304313099041533, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 16657 + }, + { + "epoch": 13.305111821086262, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 16658 + }, + { + "epoch": 13.30591054313099, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 16659 + }, + { + "epoch": 13.30670926517572, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 16660 + }, + { + "epoch": 13.307507987220447, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 16661 + }, + { + "epoch": 13.308306709265176, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 16662 + }, + { + "epoch": 13.309105431309904, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 16663 + }, + { + "epoch": 13.309904153354633, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 16664 + }, + { + "epoch": 13.310702875399361, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 16665 + }, + { + "epoch": 13.311501597444089, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 16666 + }, + { + "epoch": 13.312300319488818, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 16667 + }, + { + "epoch": 13.313099041533546, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 16668 + }, + { + "epoch": 13.313897763578275, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 16669 + }, + { + "epoch": 13.314696485623003, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 16670 + }, + { + "epoch": 13.315495207667732, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 16671 + }, + { + "epoch": 13.31629392971246, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 16672 + }, + { + "epoch": 13.31709265175719, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 16673 + }, + { + "epoch": 13.317891373801917, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 16674 + }, + { + "epoch": 13.318690095846645, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 16675 + }, + { + "epoch": 13.319488817891374, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 16676 + }, + { + "epoch": 13.320287539936102, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16677 + }, + { + "epoch": 13.321086261980831, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 16678 + }, + { + "epoch": 13.321884984025559, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 16679 + }, + { + "epoch": 13.322683706070288, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 16680 + }, + { + "epoch": 13.323482428115016, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 16681 + }, + { + "epoch": 13.324281150159745, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 16682 + }, + { + "epoch": 13.325079872204473, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 16683 + }, + { + "epoch": 13.3258785942492, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 16684 + }, + { + "epoch": 13.32667731629393, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 16685 + }, + { + "epoch": 13.327476038338657, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 16686 + }, + { + "epoch": 13.328274760383387, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 16687 + }, + { + "epoch": 13.329073482428115, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 16688 + }, + { + "epoch": 13.329872204472844, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 16689 + }, + { + "epoch": 13.330670926517572, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 16690 + }, + { + "epoch": 13.331469648562301, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 16691 + }, + { + "epoch": 13.332268370607029, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 16692 + }, + { + "epoch": 13.333067092651758, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 16693 + }, + { + "epoch": 13.333865814696486, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 16694 + }, + { + "epoch": 13.334664536741213, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 16695 + }, + { + "epoch": 13.335463258785943, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 16696 + }, + { + "epoch": 13.33626198083067, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 16697 + }, + { + "epoch": 13.3370607028754, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 16698 + }, + { + "epoch": 13.337859424920127, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 16699 + }, + { + "epoch": 13.338658146964857, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 16700 + }, + { + "epoch": 13.339456869009584, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16701 + }, + { + "epoch": 13.340255591054314, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 16702 + }, + { + "epoch": 13.341054313099042, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 16703 + }, + { + "epoch": 13.34185303514377, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 16704 + }, + { + "epoch": 13.342651757188499, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 16705 + }, + { + "epoch": 13.343450479233226, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 16706 + }, + { + "epoch": 13.344249201277956, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 16707 + }, + { + "epoch": 13.345047923322683, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 16708 + }, + { + "epoch": 13.345846645367413, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 16709 + }, + { + "epoch": 13.34664536741214, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 16710 + }, + { + "epoch": 13.34744408945687, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 16711 + }, + { + "epoch": 13.348242811501597, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 16712 + }, + { + "epoch": 13.349041533546325, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 16713 + }, + { + "epoch": 13.349840255591054, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 16714 + }, + { + "epoch": 13.350638977635782, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 16715 + }, + { + "epoch": 13.351437699680512, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 16716 + }, + { + "epoch": 13.35223642172524, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 16717 + }, + { + "epoch": 13.353035143769969, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 16718 + }, + { + "epoch": 13.353833865814696, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16719 + }, + { + "epoch": 13.354632587859426, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 16720 + }, + { + "epoch": 13.355431309904153, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16721 + }, + { + "epoch": 13.356230031948883, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 16722 + }, + { + "epoch": 13.35702875399361, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 16723 + }, + { + "epoch": 13.357827476038338, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 16724 + }, + { + "epoch": 13.358626198083067, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 16725 + }, + { + "epoch": 13.359424920127795, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 16726 + }, + { + "epoch": 13.360223642172524, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 16727 + }, + { + "epoch": 13.361022364217252, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 16728 + }, + { + "epoch": 13.361821086261982, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 16729 + }, + { + "epoch": 13.36261980830671, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 16730 + }, + { + "epoch": 13.363418530351439, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 16731 + }, + { + "epoch": 13.364217252396166, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 16732 + }, + { + "epoch": 13.365015974440894, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16733 + }, + { + "epoch": 13.365814696485623, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 16734 + }, + { + "epoch": 13.366613418530351, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 16735 + }, + { + "epoch": 13.36741214057508, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 16736 + }, + { + "epoch": 13.368210862619808, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1096, + "step": 16737 + }, + { + "epoch": 13.369009584664537, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16738 + }, + { + "epoch": 13.369808306709265, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 16739 + }, + { + "epoch": 13.370607028753994, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 16740 + }, + { + "epoch": 13.371405750798722, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 16741 + }, + { + "epoch": 13.37220447284345, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 16742 + }, + { + "epoch": 13.37300319488818, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16743 + }, + { + "epoch": 13.373801916932907, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 16744 + }, + { + "epoch": 13.374600638977636, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 16745 + }, + { + "epoch": 13.375399361022364, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 16746 + }, + { + "epoch": 13.376198083067093, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 16747 + }, + { + "epoch": 13.37699680511182, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 16748 + }, + { + "epoch": 13.37779552715655, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 16749 + }, + { + "epoch": 13.378594249201278, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 16750 + }, + { + "epoch": 13.379392971246006, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 16751 + }, + { + "epoch": 13.380191693290735, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16752 + }, + { + "epoch": 13.380990415335463, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 16753 + }, + { + "epoch": 13.381789137380192, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 16754 + }, + { + "epoch": 13.38258785942492, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 16755 + }, + { + "epoch": 13.383386581469649, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 16756 + }, + { + "epoch": 13.384185303514377, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 16757 + }, + { + "epoch": 13.384984025559106, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 16758 + }, + { + "epoch": 13.385782747603834, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 16759 + }, + { + "epoch": 13.386581469648561, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 16760 + }, + { + "epoch": 13.38738019169329, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16761 + }, + { + "epoch": 13.388178913738018, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 16762 + }, + { + "epoch": 13.388977635782748, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 16763 + }, + { + "epoch": 13.389776357827476, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 16764 + }, + { + "epoch": 13.390575079872205, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 16765 + }, + { + "epoch": 13.391373801916933, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 16766 + }, + { + "epoch": 13.392172523961662, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 16767 + }, + { + "epoch": 13.39297124600639, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 16768 + }, + { + "epoch": 13.393769968051117, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 16769 + }, + { + "epoch": 13.394568690095847, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16770 + }, + { + "epoch": 13.395367412140574, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 16771 + }, + { + "epoch": 13.396166134185304, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 16772 + }, + { + "epoch": 13.396964856230031, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 16773 + }, + { + "epoch": 13.39776357827476, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 16774 + }, + { + "epoch": 13.398562300319488, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16775 + }, + { + "epoch": 13.399361022364218, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 16776 + }, + { + "epoch": 13.400159744408946, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 16777 + }, + { + "epoch": 13.400958466453675, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 16778 + }, + { + "epoch": 13.401757188498403, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 16779 + }, + { + "epoch": 13.40255591054313, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 16780 + }, + { + "epoch": 13.40335463258786, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 16781 + }, + { + "epoch": 13.404153354632587, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 16782 + }, + { + "epoch": 13.404952076677317, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 16783 + }, + { + "epoch": 13.405750798722044, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 16784 + }, + { + "epoch": 13.406549520766774, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 16785 + }, + { + "epoch": 13.407348242811501, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 16786 + }, + { + "epoch": 13.40814696485623, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 16787 + }, + { + "epoch": 13.408945686900958, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 16788 + }, + { + "epoch": 13.409744408945686, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 16789 + }, + { + "epoch": 13.410543130990416, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 16790 + }, + { + "epoch": 13.411341853035143, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 16791 + }, + { + "epoch": 13.412140575079873, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 16792 + }, + { + "epoch": 13.4129392971246, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 16793 + }, + { + "epoch": 13.41373801916933, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 16794 + }, + { + "epoch": 13.414536741214057, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 16795 + }, + { + "epoch": 13.415335463258787, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 16796 + }, + { + "epoch": 13.416134185303514, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 16797 + }, + { + "epoch": 13.416932907348242, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 16798 + }, + { + "epoch": 13.417731629392971, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 16799 + }, + { + "epoch": 13.418530351437699, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 16800 + }, + { + "epoch": 13.419329073482428, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 16801 + }, + { + "epoch": 13.420127795527156, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 16802 + }, + { + "epoch": 13.420926517571885, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 16803 + }, + { + "epoch": 13.421725239616613, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 16804 + }, + { + "epoch": 13.422523961661343, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 16805 + }, + { + "epoch": 13.42332268370607, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 16806 + }, + { + "epoch": 13.4241214057508, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 16807 + }, + { + "epoch": 13.424920127795527, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 16808 + }, + { + "epoch": 13.425718849840255, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 16809 + }, + { + "epoch": 13.426517571884984, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 16810 + }, + { + "epoch": 13.427316293929712, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16811 + }, + { + "epoch": 13.428115015974441, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 16812 + }, + { + "epoch": 13.428913738019169, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 16813 + }, + { + "epoch": 13.429712460063898, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 16814 + }, + { + "epoch": 13.430511182108626, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 16815 + }, + { + "epoch": 13.431309904153355, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 16816 + }, + { + "epoch": 13.432108626198083, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 16817 + }, + { + "epoch": 13.43290734824281, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 16818 + }, + { + "epoch": 13.43370607028754, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 16819 + }, + { + "epoch": 13.434504792332268, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 16820 + }, + { + "epoch": 13.435303514376997, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 16821 + }, + { + "epoch": 13.436102236421725, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 16822 + }, + { + "epoch": 13.436900958466454, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 16823 + }, + { + "epoch": 13.437699680511182, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 16824 + }, + { + "epoch": 13.438498402555911, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1081, + "step": 16825 + }, + { + "epoch": 13.439297124600639, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 16826 + }, + { + "epoch": 13.440095846645367, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16827 + }, + { + "epoch": 13.440894568690096, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 16828 + }, + { + "epoch": 13.441693290734824, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 16829 + }, + { + "epoch": 13.442492012779553, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 16830 + }, + { + "epoch": 13.44329073482428, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 16831 + }, + { + "epoch": 13.44408945686901, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 16832 + }, + { + "epoch": 13.444888178913738, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 16833 + }, + { + "epoch": 13.445686900958467, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 16834 + }, + { + "epoch": 13.446485623003195, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 16835 + }, + { + "epoch": 13.447284345047922, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 16836 + }, + { + "epoch": 13.448083067092652, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 16837 + }, + { + "epoch": 13.44888178913738, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 16838 + }, + { + "epoch": 13.449680511182109, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 16839 + }, + { + "epoch": 13.450479233226837, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 16840 + }, + { + "epoch": 13.451277955271566, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 16841 + }, + { + "epoch": 13.452076677316294, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 16842 + }, + { + "epoch": 13.452875399361023, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 16843 + }, + { + "epoch": 13.45367412140575, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 16844 + }, + { + "epoch": 13.454472843450478, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 16845 + }, + { + "epoch": 13.455271565495208, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 16846 + }, + { + "epoch": 13.456070287539935, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 16847 + }, + { + "epoch": 13.456869009584665, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 16848 + }, + { + "epoch": 13.457667731629392, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 16849 + }, + { + "epoch": 13.458466453674122, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 16850 + }, + { + "epoch": 13.45926517571885, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16851 + }, + { + "epoch": 13.460063897763579, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 16852 + }, + { + "epoch": 13.460862619808307, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 16853 + }, + { + "epoch": 13.461661341853036, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 16854 + }, + { + "epoch": 13.462460063897764, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 16855 + }, + { + "epoch": 13.463258785942491, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 16856 + }, + { + "epoch": 13.46405750798722, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 16857 + }, + { + "epoch": 13.464856230031948, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 16858 + }, + { + "epoch": 13.465654952076678, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 16859 + }, + { + "epoch": 13.466453674121405, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 16860 + }, + { + "epoch": 13.467252396166135, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 16861 + }, + { + "epoch": 13.468051118210862, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 16862 + }, + { + "epoch": 13.468849840255592, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 16863 + }, + { + "epoch": 13.46964856230032, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 16864 + }, + { + "epoch": 13.470447284345047, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 16865 + }, + { + "epoch": 13.471246006389777, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 16866 + }, + { + "epoch": 13.472044728434504, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16867 + }, + { + "epoch": 13.472843450479234, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 16868 + }, + { + "epoch": 13.473642172523961, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 16869 + }, + { + "epoch": 13.47444089456869, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 16870 + }, + { + "epoch": 13.475239616613418, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 16871 + }, + { + "epoch": 13.476038338658148, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 16872 + }, + { + "epoch": 13.476837060702875, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 16873 + }, + { + "epoch": 13.477635782747603, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 16874 + }, + { + "epoch": 13.478434504792332, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 16875 + }, + { + "epoch": 13.47923322683706, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 16876 + }, + { + "epoch": 13.48003194888179, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 16877 + }, + { + "epoch": 13.480830670926517, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 16878 + }, + { + "epoch": 13.481629392971247, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 16879 + }, + { + "epoch": 13.482428115015974, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 16880 + }, + { + "epoch": 13.483226837060704, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 16881 + }, + { + "epoch": 13.484025559105431, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 16882 + }, + { + "epoch": 13.48482428115016, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 16883 + }, + { + "epoch": 13.485623003194888, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 16884 + }, + { + "epoch": 13.486421725239616, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 16885 + }, + { + "epoch": 13.487220447284345, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 16886 + }, + { + "epoch": 13.488019169329073, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 16887 + }, + { + "epoch": 13.488817891373802, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 16888 + }, + { + "epoch": 13.48961661341853, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 16889 + }, + { + "epoch": 13.49041533546326, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1106, + "step": 16890 + }, + { + "epoch": 13.491214057507987, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 16891 + }, + { + "epoch": 13.492012779552716, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 16892 + }, + { + "epoch": 13.492811501597444, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 16893 + }, + { + "epoch": 13.493610223642172, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 16894 + }, + { + "epoch": 13.494408945686901, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 16895 + }, + { + "epoch": 13.495207667731629, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 16896 + }, + { + "epoch": 13.496006389776358, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 16897 + }, + { + "epoch": 13.496805111821086, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 16898 + }, + { + "epoch": 13.497603833865815, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 16899 + }, + { + "epoch": 13.498402555910543, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 16900 + }, + { + "epoch": 13.499201277955272, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 16901 + }, + { + "epoch": 13.5, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 16902 + }, + { + "epoch": 13.500798722044728, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 16903 + }, + { + "epoch": 13.501597444089457, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 16904 + }, + { + "epoch": 13.502396166134185, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 16905 + }, + { + "epoch": 13.503194888178914, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 16906 + }, + { + "epoch": 13.503993610223642, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 16907 + }, + { + "epoch": 13.504792332268371, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16908 + }, + { + "epoch": 13.505591054313099, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 16909 + }, + { + "epoch": 13.506389776357828, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 16910 + }, + { + "epoch": 13.507188498402556, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 16911 + }, + { + "epoch": 13.507987220447284, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 16912 + }, + { + "epoch": 13.508785942492013, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 16913 + }, + { + "epoch": 13.50958466453674, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 16914 + }, + { + "epoch": 13.51038338658147, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 16915 + }, + { + "epoch": 13.511182108626198, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 16916 + }, + { + "epoch": 13.511980830670927, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 16917 + }, + { + "epoch": 13.512779552715655, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 16918 + }, + { + "epoch": 13.513578274760384, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 16919 + }, + { + "epoch": 13.514376996805112, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 16920 + }, + { + "epoch": 13.51517571884984, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 16921 + }, + { + "epoch": 13.515974440894569, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 16922 + }, + { + "epoch": 13.516773162939296, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 16923 + }, + { + "epoch": 13.517571884984026, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 16924 + }, + { + "epoch": 13.518370607028753, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 16925 + }, + { + "epoch": 13.519169329073483, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 16926 + }, + { + "epoch": 13.51996805111821, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 16927 + }, + { + "epoch": 13.52076677316294, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 16928 + }, + { + "epoch": 13.521565495207668, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 16929 + }, + { + "epoch": 13.522364217252395, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 16930 + }, + { + "epoch": 13.523162939297125, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 16931 + }, + { + "epoch": 13.523961661341852, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 16932 + }, + { + "epoch": 13.524760383386582, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 16933 + }, + { + "epoch": 13.52555910543131, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 16934 + }, + { + "epoch": 13.526357827476039, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 16935 + }, + { + "epoch": 13.527156549520766, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 16936 + }, + { + "epoch": 13.527955271565496, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 16937 + }, + { + "epoch": 13.528753993610223, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 16938 + }, + { + "epoch": 13.529552715654953, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 16939 + }, + { + "epoch": 13.53035143769968, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 16940 + }, + { + "epoch": 13.531150159744408, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 16941 + }, + { + "epoch": 13.531948881789138, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 16942 + }, + { + "epoch": 13.532747603833865, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 16943 + }, + { + "epoch": 13.533546325878595, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 16944 + }, + { + "epoch": 13.534345047923322, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 16945 + }, + { + "epoch": 13.535143769968052, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 16946 + }, + { + "epoch": 13.53594249201278, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 16947 + }, + { + "epoch": 13.536741214057509, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 16948 + }, + { + "epoch": 13.537539936102236, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 16949 + }, + { + "epoch": 13.538338658146966, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 16950 + }, + { + "epoch": 13.539137380191693, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 16951 + }, + { + "epoch": 13.539936102236421, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 16952 + }, + { + "epoch": 13.54073482428115, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 16953 + }, + { + "epoch": 13.541533546325878, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 16954 + }, + { + "epoch": 13.542332268370608, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 16955 + }, + { + "epoch": 13.543130990415335, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 16956 + }, + { + "epoch": 13.543929712460065, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 16957 + }, + { + "epoch": 13.544728434504792, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16958 + }, + { + "epoch": 13.545527156549522, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 16959 + }, + { + "epoch": 13.54632587859425, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 16960 + }, + { + "epoch": 13.547124600638977, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 16961 + }, + { + "epoch": 13.547923322683706, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 16962 + }, + { + "epoch": 13.548722044728434, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 16963 + }, + { + "epoch": 13.549520766773163, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 16964 + }, + { + "epoch": 13.550319488817891, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 16965 + }, + { + "epoch": 13.55111821086262, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16966 + }, + { + "epoch": 13.551916932907348, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 16967 + }, + { + "epoch": 13.552715654952078, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 16968 + }, + { + "epoch": 13.553514376996805, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 16969 + }, + { + "epoch": 13.554313099041533, + "grad_norm": 0.34765625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 16970 + }, + { + "epoch": 13.555111821086262, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 16971 + }, + { + "epoch": 13.55591054313099, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 16972 + }, + { + "epoch": 13.55670926517572, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 16973 + }, + { + "epoch": 13.557507987220447, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 16974 + }, + { + "epoch": 13.558306709265176, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 16975 + }, + { + "epoch": 13.559105431309904, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 16976 + }, + { + "epoch": 13.559904153354633, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 16977 + }, + { + "epoch": 13.560702875399361, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 16978 + }, + { + "epoch": 13.561501597444089, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 16979 + }, + { + "epoch": 13.562300319488818, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 16980 + }, + { + "epoch": 13.563099041533546, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 16981 + }, + { + "epoch": 13.563897763578275, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 16982 + }, + { + "epoch": 13.564696485623003, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 16983 + }, + { + "epoch": 13.565495207667732, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 16984 + }, + { + "epoch": 13.56629392971246, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 16985 + }, + { + "epoch": 13.56709265175719, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 16986 + }, + { + "epoch": 13.567891373801917, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 16987 + }, + { + "epoch": 13.568690095846645, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 16988 + }, + { + "epoch": 13.569488817891374, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 16989 + }, + { + "epoch": 13.570287539936102, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 16990 + }, + { + "epoch": 13.571086261980831, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 16991 + }, + { + "epoch": 13.571884984025559, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 16992 + }, + { + "epoch": 13.572683706070288, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 16993 + }, + { + "epoch": 13.573482428115016, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 16994 + }, + { + "epoch": 13.574281150159745, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 16995 + }, + { + "epoch": 13.575079872204473, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 16996 + }, + { + "epoch": 13.5758785942492, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 16997 + }, + { + "epoch": 13.57667731629393, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 16998 + }, + { + "epoch": 13.577476038338657, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 16999 + }, + { + "epoch": 13.578274760383387, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 17000 + }, + { + "epoch": 13.579073482428115, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 17001 + }, + { + "epoch": 13.579872204472844, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 17002 + }, + { + "epoch": 13.580670926517572, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 17003 + }, + { + "epoch": 13.581469648562301, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 17004 + }, + { + "epoch": 13.582268370607029, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 17005 + }, + { + "epoch": 13.583067092651756, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17006 + }, + { + "epoch": 13.583865814696486, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 17007 + }, + { + "epoch": 13.584664536741213, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 17008 + }, + { + "epoch": 13.585463258785943, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17009 + }, + { + "epoch": 13.58626198083067, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 17010 + }, + { + "epoch": 13.5870607028754, + "grad_norm": 0.318359375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 17011 + }, + { + "epoch": 13.587859424920127, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 17012 + }, + { + "epoch": 13.588658146964857, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 17013 + }, + { + "epoch": 13.589456869009584, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 17014 + }, + { + "epoch": 13.590255591054314, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 17015 + }, + { + "epoch": 13.591054313099042, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 17016 + }, + { + "epoch": 13.59185303514377, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 17017 + }, + { + "epoch": 13.592651757188499, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 17018 + }, + { + "epoch": 13.593450479233226, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 17019 + }, + { + "epoch": 13.594249201277956, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 17020 + }, + { + "epoch": 13.595047923322683, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 17021 + }, + { + "epoch": 13.595846645367413, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17022 + }, + { + "epoch": 13.59664536741214, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 17023 + }, + { + "epoch": 13.59744408945687, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 17024 + }, + { + "epoch": 13.598242811501597, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 17025 + }, + { + "epoch": 13.599041533546325, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 17026 + }, + { + "epoch": 13.599840255591054, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 17027 + }, + { + "epoch": 13.600638977635782, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 17028 + }, + { + "epoch": 13.601437699680512, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 17029 + }, + { + "epoch": 13.60223642172524, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 17030 + }, + { + "epoch": 13.603035143769969, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 17031 + }, + { + "epoch": 13.603833865814696, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 17032 + }, + { + "epoch": 13.604632587859426, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 17033 + }, + { + "epoch": 13.605431309904153, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 17034 + }, + { + "epoch": 13.606230031948883, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 17035 + }, + { + "epoch": 13.60702875399361, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 17036 + }, + { + "epoch": 13.607827476038338, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 17037 + }, + { + "epoch": 13.608626198083067, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 17038 + }, + { + "epoch": 13.609424920127795, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 17039 + }, + { + "epoch": 13.610223642172524, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 17040 + }, + { + "epoch": 13.611022364217252, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 17041 + }, + { + "epoch": 13.611821086261982, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 17042 + }, + { + "epoch": 13.61261980830671, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 17043 + }, + { + "epoch": 13.613418530351439, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 17044 + }, + { + "epoch": 13.614217252396166, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 17045 + }, + { + "epoch": 13.615015974440894, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 17046 + }, + { + "epoch": 13.615814696485623, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 17047 + }, + { + "epoch": 13.616613418530351, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 17048 + }, + { + "epoch": 13.61741214057508, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 17049 + }, + { + "epoch": 13.618210862619808, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17050 + }, + { + "epoch": 13.619009584664537, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 17051 + }, + { + "epoch": 13.619808306709265, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 17052 + }, + { + "epoch": 13.620607028753994, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 17053 + }, + { + "epoch": 13.621405750798722, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 17054 + }, + { + "epoch": 13.62220447284345, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 17055 + }, + { + "epoch": 13.62300319488818, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 17056 + }, + { + "epoch": 13.623801916932907, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 17057 + }, + { + "epoch": 13.624600638977636, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 17058 + }, + { + "epoch": 13.625399361022364, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 17059 + }, + { + "epoch": 13.626198083067093, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 17060 + }, + { + "epoch": 13.62699680511182, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 17061 + }, + { + "epoch": 13.62779552715655, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 17062 + }, + { + "epoch": 13.628594249201278, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 17063 + }, + { + "epoch": 13.629392971246006, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17064 + }, + { + "epoch": 13.630191693290735, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 17065 + }, + { + "epoch": 13.630990415335463, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 17066 + }, + { + "epoch": 13.631789137380192, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 17067 + }, + { + "epoch": 13.63258785942492, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 17068 + }, + { + "epoch": 13.633386581469649, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 17069 + }, + { + "epoch": 13.634185303514377, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 17070 + }, + { + "epoch": 13.634984025559106, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 17071 + }, + { + "epoch": 13.635782747603834, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 17072 + }, + { + "epoch": 13.636581469648561, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17073 + }, + { + "epoch": 13.63738019169329, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 17074 + }, + { + "epoch": 13.638178913738018, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 17075 + }, + { + "epoch": 13.638977635782748, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 17076 + }, + { + "epoch": 13.639776357827476, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 17077 + }, + { + "epoch": 13.640575079872205, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 17078 + }, + { + "epoch": 13.641373801916933, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 17079 + }, + { + "epoch": 13.642172523961662, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 17080 + }, + { + "epoch": 13.64297124600639, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 17081 + }, + { + "epoch": 13.643769968051117, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 17082 + }, + { + "epoch": 13.644568690095847, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17083 + }, + { + "epoch": 13.645367412140574, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 17084 + }, + { + "epoch": 13.646166134185304, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 17085 + }, + { + "epoch": 13.646964856230031, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 17086 + }, + { + "epoch": 13.64776357827476, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 17087 + }, + { + "epoch": 13.648562300319488, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 17088 + }, + { + "epoch": 13.649361022364218, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 17089 + }, + { + "epoch": 13.650159744408946, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 17090 + }, + { + "epoch": 13.650958466453675, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 17091 + }, + { + "epoch": 13.651757188498403, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 17092 + }, + { + "epoch": 13.65255591054313, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 17093 + }, + { + "epoch": 13.65335463258786, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 17094 + }, + { + "epoch": 13.654153354632587, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 17095 + }, + { + "epoch": 13.654952076677317, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 17096 + }, + { + "epoch": 13.655750798722044, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 17097 + }, + { + "epoch": 13.656549520766774, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 17098 + }, + { + "epoch": 13.657348242811501, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 17099 + }, + { + "epoch": 13.65814696485623, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 17100 + }, + { + "epoch": 13.658945686900958, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 17101 + }, + { + "epoch": 13.659744408945686, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 17102 + }, + { + "epoch": 13.660543130990416, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 17103 + }, + { + "epoch": 13.661341853035143, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 17104 + }, + { + "epoch": 13.662140575079873, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 17105 + }, + { + "epoch": 13.6629392971246, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 17106 + }, + { + "epoch": 13.66373801916933, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 17107 + }, + { + "epoch": 13.664536741214057, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 17108 + }, + { + "epoch": 13.665335463258787, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 17109 + }, + { + "epoch": 13.666134185303514, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 17110 + }, + { + "epoch": 13.666932907348244, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 17111 + }, + { + "epoch": 13.667731629392971, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17112 + }, + { + "epoch": 13.668530351437699, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 17113 + }, + { + "epoch": 13.669329073482428, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 17114 + }, + { + "epoch": 13.670127795527156, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17115 + }, + { + "epoch": 13.670926517571885, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 17116 + }, + { + "epoch": 13.671725239616613, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 17117 + }, + { + "epoch": 13.672523961661343, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 17118 + }, + { + "epoch": 13.67332268370607, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 17119 + }, + { + "epoch": 13.6741214057508, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 17120 + }, + { + "epoch": 13.674920127795527, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 17121 + }, + { + "epoch": 13.675718849840255, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 17122 + }, + { + "epoch": 13.676517571884984, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 17123 + }, + { + "epoch": 13.677316293929712, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17124 + }, + { + "epoch": 13.678115015974441, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 17125 + }, + { + "epoch": 13.678913738019169, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17126 + }, + { + "epoch": 13.679712460063898, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 17127 + }, + { + "epoch": 13.680511182108626, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 17128 + }, + { + "epoch": 13.681309904153355, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 17129 + }, + { + "epoch": 13.682108626198083, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 17130 + }, + { + "epoch": 13.68290734824281, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17131 + }, + { + "epoch": 13.68370607028754, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 17132 + }, + { + "epoch": 13.684504792332268, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 17133 + }, + { + "epoch": 13.685303514376997, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 17134 + }, + { + "epoch": 13.686102236421725, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 17135 + }, + { + "epoch": 13.686900958466454, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 17136 + }, + { + "epoch": 13.687699680511182, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17137 + }, + { + "epoch": 13.688498402555911, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 17138 + }, + { + "epoch": 13.689297124600639, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 17139 + }, + { + "epoch": 13.690095846645367, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 17140 + }, + { + "epoch": 13.690894568690096, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 17141 + }, + { + "epoch": 13.691693290734824, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 17142 + }, + { + "epoch": 13.692492012779553, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 17143 + }, + { + "epoch": 13.69329073482428, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 17144 + }, + { + "epoch": 13.69408945686901, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 17145 + }, + { + "epoch": 13.694888178913738, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 17146 + }, + { + "epoch": 13.695686900958467, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 17147 + }, + { + "epoch": 13.696485623003195, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 17148 + }, + { + "epoch": 13.697284345047922, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 17149 + }, + { + "epoch": 13.698083067092652, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17150 + }, + { + "epoch": 13.69888178913738, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1087, + "step": 17151 + }, + { + "epoch": 13.699680511182109, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 17152 + }, + { + "epoch": 13.700479233226837, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 17153 + }, + { + "epoch": 13.701277955271566, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1099, + "step": 17154 + }, + { + "epoch": 13.702076677316294, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 17155 + }, + { + "epoch": 13.702875399361023, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17156 + }, + { + "epoch": 13.70367412140575, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17157 + }, + { + "epoch": 13.704472843450478, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 17158 + }, + { + "epoch": 13.705271565495208, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 17159 + }, + { + "epoch": 13.706070287539935, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 17160 + }, + { + "epoch": 13.706869009584665, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 17161 + }, + { + "epoch": 13.707667731629392, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 17162 + }, + { + "epoch": 13.708466453674122, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 17163 + }, + { + "epoch": 13.70926517571885, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 17164 + }, + { + "epoch": 13.710063897763579, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 17165 + }, + { + "epoch": 13.710862619808307, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 17166 + }, + { + "epoch": 13.711661341853034, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 17167 + }, + { + "epoch": 13.712460063897764, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17168 + }, + { + "epoch": 13.713258785942491, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 17169 + }, + { + "epoch": 13.71405750798722, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 17170 + }, + { + "epoch": 13.714856230031948, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 17171 + }, + { + "epoch": 13.715654952076678, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 17172 + }, + { + "epoch": 13.716453674121405, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 17173 + }, + { + "epoch": 13.717252396166135, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 17174 + }, + { + "epoch": 13.718051118210862, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 17175 + }, + { + "epoch": 13.718849840255592, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 17176 + }, + { + "epoch": 13.71964856230032, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 17177 + }, + { + "epoch": 13.720447284345047, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1098, + "step": 17178 + }, + { + "epoch": 13.721246006389777, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 17179 + }, + { + "epoch": 13.722044728434504, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 17180 + }, + { + "epoch": 13.722843450479234, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 17181 + }, + { + "epoch": 13.723642172523961, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 17182 + }, + { + "epoch": 13.72444089456869, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 17183 + }, + { + "epoch": 13.725239616613418, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 17184 + }, + { + "epoch": 13.726038338658148, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 17185 + }, + { + "epoch": 13.726837060702875, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 17186 + }, + { + "epoch": 13.727635782747605, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17187 + }, + { + "epoch": 13.728434504792332, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 17188 + }, + { + "epoch": 13.72923322683706, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 17189 + }, + { + "epoch": 13.73003194888179, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 17190 + }, + { + "epoch": 13.730830670926517, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 17191 + }, + { + "epoch": 13.731629392971247, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 17192 + }, + { + "epoch": 13.732428115015974, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17193 + }, + { + "epoch": 13.733226837060704, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 17194 + }, + { + "epoch": 13.734025559105431, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 17195 + }, + { + "epoch": 13.73482428115016, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 17196 + }, + { + "epoch": 13.735623003194888, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 17197 + }, + { + "epoch": 13.736421725239616, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 17198 + }, + { + "epoch": 13.737220447284345, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 17199 + }, + { + "epoch": 13.738019169329073, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 17200 + }, + { + "epoch": 13.738817891373802, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 17201 + }, + { + "epoch": 13.73961661341853, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 17202 + }, + { + "epoch": 13.74041533546326, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 17203 + }, + { + "epoch": 13.741214057507987, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 17204 + }, + { + "epoch": 13.742012779552716, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 17205 + }, + { + "epoch": 13.742811501597444, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 17206 + }, + { + "epoch": 13.743610223642172, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 17207 + }, + { + "epoch": 13.744408945686901, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 17208 + }, + { + "epoch": 13.745207667731629, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 17209 + }, + { + "epoch": 13.746006389776358, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 17210 + }, + { + "epoch": 13.746805111821086, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17211 + }, + { + "epoch": 13.747603833865815, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 17212 + }, + { + "epoch": 13.748402555910543, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 17213 + }, + { + "epoch": 13.749201277955272, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 17214 + }, + { + "epoch": 13.75, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17215 + }, + { + "epoch": 13.750798722044728, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 17216 + }, + { + "epoch": 13.751597444089457, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 17217 + }, + { + "epoch": 13.752396166134185, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 17218 + }, + { + "epoch": 13.753194888178914, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 17219 + }, + { + "epoch": 13.753993610223642, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 17220 + }, + { + "epoch": 13.754792332268371, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17221 + }, + { + "epoch": 13.755591054313099, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 17222 + }, + { + "epoch": 13.756389776357828, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 17223 + }, + { + "epoch": 13.757188498402556, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 17224 + }, + { + "epoch": 13.757987220447284, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 17225 + }, + { + "epoch": 13.758785942492013, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 17226 + }, + { + "epoch": 13.75958466453674, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 17227 + }, + { + "epoch": 13.76038338658147, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 17228 + }, + { + "epoch": 13.761182108626198, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 17229 + }, + { + "epoch": 13.761980830670927, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 17230 + }, + { + "epoch": 13.762779552715655, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17231 + }, + { + "epoch": 13.763578274760384, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 17232 + }, + { + "epoch": 13.764376996805112, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17233 + }, + { + "epoch": 13.76517571884984, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 17234 + }, + { + "epoch": 13.765974440894569, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 17235 + }, + { + "epoch": 13.766773162939296, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 17236 + }, + { + "epoch": 13.767571884984026, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 17237 + }, + { + "epoch": 13.768370607028753, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 17238 + }, + { + "epoch": 13.769169329073483, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 17239 + }, + { + "epoch": 13.76996805111821, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 17240 + }, + { + "epoch": 13.77076677316294, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17241 + }, + { + "epoch": 13.771565495207668, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 17242 + }, + { + "epoch": 13.772364217252395, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 17243 + }, + { + "epoch": 13.773162939297125, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 17244 + }, + { + "epoch": 13.773961661341852, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 17245 + }, + { + "epoch": 13.774760383386582, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 17246 + }, + { + "epoch": 13.77555910543131, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 17247 + }, + { + "epoch": 13.776357827476039, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 17248 + }, + { + "epoch": 13.777156549520766, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 17249 + }, + { + "epoch": 13.777955271565496, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 17250 + }, + { + "epoch": 13.778753993610223, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 17251 + }, + { + "epoch": 13.779552715654953, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 17252 + }, + { + "epoch": 13.78035143769968, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 17253 + }, + { + "epoch": 13.781150159744408, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 17254 + }, + { + "epoch": 13.781948881789138, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 17255 + }, + { + "epoch": 13.782747603833865, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 17256 + }, + { + "epoch": 13.783546325878595, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 17257 + }, + { + "epoch": 13.784345047923322, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 17258 + }, + { + "epoch": 13.785143769968052, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 17259 + }, + { + "epoch": 13.78594249201278, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 17260 + }, + { + "epoch": 13.786741214057509, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 17261 + }, + { + "epoch": 13.787539936102236, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 17262 + }, + { + "epoch": 13.788338658146966, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 17263 + }, + { + "epoch": 13.789137380191693, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 17264 + }, + { + "epoch": 13.789936102236421, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 17265 + }, + { + "epoch": 13.79073482428115, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 17266 + }, + { + "epoch": 13.791533546325878, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 17267 + }, + { + "epoch": 13.792332268370608, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 17268 + }, + { + "epoch": 13.793130990415335, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 17269 + }, + { + "epoch": 13.793929712460065, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 17270 + }, + { + "epoch": 13.794728434504792, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17271 + }, + { + "epoch": 13.795527156549522, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 17272 + }, + { + "epoch": 13.79632587859425, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 17273 + }, + { + "epoch": 13.797124600638977, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 17274 + }, + { + "epoch": 13.797923322683706, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 17275 + }, + { + "epoch": 13.798722044728434, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 17276 + }, + { + "epoch": 13.799520766773163, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 17277 + }, + { + "epoch": 13.800319488817891, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 17278 + }, + { + "epoch": 13.80111821086262, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 17279 + }, + { + "epoch": 13.801916932907348, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 17280 + }, + { + "epoch": 13.802715654952078, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 17281 + }, + { + "epoch": 13.803514376996805, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 17282 + }, + { + "epoch": 13.804313099041533, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17283 + }, + { + "epoch": 13.805111821086262, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 17284 + }, + { + "epoch": 13.80591054313099, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 17285 + }, + { + "epoch": 13.80670926517572, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 17286 + }, + { + "epoch": 13.807507987220447, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 17287 + }, + { + "epoch": 13.808306709265176, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 17288 + }, + { + "epoch": 13.809105431309904, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 17289 + }, + { + "epoch": 13.809904153354633, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 17290 + }, + { + "epoch": 13.810702875399361, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 17291 + }, + { + "epoch": 13.811501597444089, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 17292 + }, + { + "epoch": 13.812300319488818, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 17293 + }, + { + "epoch": 13.813099041533546, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 17294 + }, + { + "epoch": 13.813897763578275, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 17295 + }, + { + "epoch": 13.814696485623003, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 17296 + }, + { + "epoch": 13.815495207667732, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 17297 + }, + { + "epoch": 13.81629392971246, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 17298 + }, + { + "epoch": 13.81709265175719, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 17299 + }, + { + "epoch": 13.817891373801917, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 17300 + }, + { + "epoch": 13.818690095846645, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17301 + }, + { + "epoch": 13.819488817891374, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 17302 + }, + { + "epoch": 13.820287539936102, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 17303 + }, + { + "epoch": 13.821086261980831, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 17304 + }, + { + "epoch": 13.821884984025559, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 17305 + }, + { + "epoch": 13.822683706070288, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17306 + }, + { + "epoch": 13.823482428115016, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 17307 + }, + { + "epoch": 13.824281150159745, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 17308 + }, + { + "epoch": 13.825079872204473, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 17309 + }, + { + "epoch": 13.8258785942492, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 17310 + }, + { + "epoch": 13.82667731629393, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 17311 + }, + { + "epoch": 13.827476038338657, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 17312 + }, + { + "epoch": 13.828274760383387, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 17313 + }, + { + "epoch": 13.829073482428115, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 17314 + }, + { + "epoch": 13.829872204472844, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 17315 + }, + { + "epoch": 13.830670926517572, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17316 + }, + { + "epoch": 13.831469648562301, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 17317 + }, + { + "epoch": 13.832268370607029, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 17318 + }, + { + "epoch": 13.833067092651756, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 17319 + }, + { + "epoch": 13.833865814696486, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 17320 + }, + { + "epoch": 13.834664536741213, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 17321 + }, + { + "epoch": 13.835463258785943, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 17322 + }, + { + "epoch": 13.83626198083067, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 17323 + }, + { + "epoch": 13.8370607028754, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 17324 + }, + { + "epoch": 13.837859424920127, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 17325 + }, + { + "epoch": 13.838658146964857, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 17326 + }, + { + "epoch": 13.839456869009584, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 17327 + }, + { + "epoch": 13.840255591054314, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 17328 + }, + { + "epoch": 13.841054313099042, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 17329 + }, + { + "epoch": 13.84185303514377, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 17330 + }, + { + "epoch": 13.842651757188499, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 17331 + }, + { + "epoch": 13.843450479233226, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 17332 + }, + { + "epoch": 13.844249201277956, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 17333 + }, + { + "epoch": 13.845047923322683, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 17334 + }, + { + "epoch": 13.845846645367413, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 17335 + }, + { + "epoch": 13.84664536741214, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 17336 + }, + { + "epoch": 13.84744408945687, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 17337 + }, + { + "epoch": 13.848242811501597, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17338 + }, + { + "epoch": 13.849041533546325, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 17339 + }, + { + "epoch": 13.849840255591054, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 17340 + }, + { + "epoch": 13.850638977635782, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 17341 + }, + { + "epoch": 13.851437699680512, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 17342 + }, + { + "epoch": 13.85223642172524, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 17343 + }, + { + "epoch": 13.853035143769969, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 17344 + }, + { + "epoch": 13.853833865814696, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 17345 + }, + { + "epoch": 13.854632587859426, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 17346 + }, + { + "epoch": 13.855431309904153, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17347 + }, + { + "epoch": 13.856230031948883, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 17348 + }, + { + "epoch": 13.85702875399361, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 17349 + }, + { + "epoch": 13.857827476038338, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 17350 + }, + { + "epoch": 13.858626198083067, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 17351 + }, + { + "epoch": 13.859424920127795, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 17352 + }, + { + "epoch": 13.860223642172524, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 17353 + }, + { + "epoch": 13.861022364217252, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 17354 + }, + { + "epoch": 13.861821086261982, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 17355 + }, + { + "epoch": 13.86261980830671, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 17356 + }, + { + "epoch": 13.863418530351439, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 17357 + }, + { + "epoch": 13.864217252396166, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 17358 + }, + { + "epoch": 13.865015974440894, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 17359 + }, + { + "epoch": 13.865814696485623, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 17360 + }, + { + "epoch": 13.866613418530351, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 17361 + }, + { + "epoch": 13.86741214057508, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 17362 + }, + { + "epoch": 13.868210862619808, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 17363 + }, + { + "epoch": 13.869009584664537, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 17364 + }, + { + "epoch": 13.869808306709265, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 17365 + }, + { + "epoch": 13.870607028753994, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 17366 + }, + { + "epoch": 13.871405750798722, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 17367 + }, + { + "epoch": 13.87220447284345, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 17368 + }, + { + "epoch": 13.87300319488818, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 17369 + }, + { + "epoch": 13.873801916932907, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 17370 + }, + { + "epoch": 13.874600638977636, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 17371 + }, + { + "epoch": 13.875399361022364, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 17372 + }, + { + "epoch": 13.876198083067093, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 17373 + }, + { + "epoch": 13.87699680511182, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 17374 + }, + { + "epoch": 13.87779552715655, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17375 + }, + { + "epoch": 13.878594249201278, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 17376 + }, + { + "epoch": 13.879392971246006, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 17377 + }, + { + "epoch": 13.880191693290735, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 17378 + }, + { + "epoch": 13.880990415335463, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 17379 + }, + { + "epoch": 13.881789137380192, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 17380 + }, + { + "epoch": 13.88258785942492, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 17381 + }, + { + "epoch": 13.883386581469649, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17382 + }, + { + "epoch": 13.884185303514377, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 17383 + }, + { + "epoch": 13.884984025559106, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17384 + }, + { + "epoch": 13.885782747603834, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 17385 + }, + { + "epoch": 13.886581469648561, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 17386 + }, + { + "epoch": 13.88738019169329, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17387 + }, + { + "epoch": 13.888178913738018, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 17388 + }, + { + "epoch": 13.888977635782748, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 17389 + }, + { + "epoch": 13.889776357827476, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 17390 + }, + { + "epoch": 13.890575079872205, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17391 + }, + { + "epoch": 13.891373801916933, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 17392 + }, + { + "epoch": 13.892172523961662, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 17393 + }, + { + "epoch": 13.89297124600639, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 17394 + }, + { + "epoch": 13.893769968051117, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 17395 + }, + { + "epoch": 13.894568690095847, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17396 + }, + { + "epoch": 13.895367412140574, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 17397 + }, + { + "epoch": 13.896166134185304, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17398 + }, + { + "epoch": 13.896964856230031, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 17399 + }, + { + "epoch": 13.89776357827476, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17400 + }, + { + "epoch": 13.898562300319488, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 17401 + }, + { + "epoch": 13.899361022364218, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 17402 + }, + { + "epoch": 13.900159744408946, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17403 + }, + { + "epoch": 13.900958466453675, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 17404 + }, + { + "epoch": 13.901757188498403, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 17405 + }, + { + "epoch": 13.90255591054313, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 17406 + }, + { + "epoch": 13.90335463258786, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 17407 + }, + { + "epoch": 13.904153354632587, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 17408 + }, + { + "epoch": 13.904952076677317, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 17409 + }, + { + "epoch": 13.905750798722044, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 17410 + }, + { + "epoch": 13.906549520766774, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 17411 + }, + { + "epoch": 13.907348242811501, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 17412 + }, + { + "epoch": 13.90814696485623, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 17413 + }, + { + "epoch": 13.908945686900958, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 17414 + }, + { + "epoch": 13.909744408945686, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 17415 + }, + { + "epoch": 13.910543130990416, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 17416 + }, + { + "epoch": 13.911341853035143, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 17417 + }, + { + "epoch": 13.912140575079873, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 17418 + }, + { + "epoch": 13.9129392971246, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 17419 + }, + { + "epoch": 13.91373801916933, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 17420 + }, + { + "epoch": 13.914536741214057, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 17421 + }, + { + "epoch": 13.915335463258787, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 17422 + }, + { + "epoch": 13.916134185303514, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 17423 + }, + { + "epoch": 13.916932907348244, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 17424 + }, + { + "epoch": 13.917731629392971, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 17425 + }, + { + "epoch": 13.918530351437699, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 17426 + }, + { + "epoch": 13.919329073482428, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 17427 + }, + { + "epoch": 13.920127795527156, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 17428 + }, + { + "epoch": 13.920926517571885, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 17429 + }, + { + "epoch": 13.921725239616613, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 17430 + }, + { + "epoch": 13.922523961661343, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 17431 + }, + { + "epoch": 13.92332268370607, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 17432 + }, + { + "epoch": 13.9241214057508, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 17433 + }, + { + "epoch": 13.924920127795527, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 17434 + }, + { + "epoch": 13.925718849840255, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 17435 + }, + { + "epoch": 13.926517571884984, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17436 + }, + { + "epoch": 13.927316293929712, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 17437 + }, + { + "epoch": 13.928115015974441, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17438 + }, + { + "epoch": 13.928913738019169, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 17439 + }, + { + "epoch": 13.929712460063898, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 17440 + }, + { + "epoch": 13.930511182108626, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17441 + }, + { + "epoch": 13.931309904153355, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 17442 + }, + { + "epoch": 13.932108626198083, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 17443 + }, + { + "epoch": 13.93290734824281, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 17444 + }, + { + "epoch": 13.93370607028754, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17445 + }, + { + "epoch": 13.934504792332268, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 17446 + }, + { + "epoch": 13.935303514376997, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 17447 + }, + { + "epoch": 13.936102236421725, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 17448 + }, + { + "epoch": 13.936900958466454, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 17449 + }, + { + "epoch": 13.937699680511182, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 17450 + }, + { + "epoch": 13.938498402555911, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 17451 + }, + { + "epoch": 13.939297124600639, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 17452 + }, + { + "epoch": 13.940095846645367, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 17453 + }, + { + "epoch": 13.940894568690096, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 17454 + }, + { + "epoch": 13.941693290734824, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 17455 + }, + { + "epoch": 13.942492012779553, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 17456 + }, + { + "epoch": 13.94329073482428, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 17457 + }, + { + "epoch": 13.94408945686901, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17458 + }, + { + "epoch": 13.944888178913738, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 17459 + }, + { + "epoch": 13.945686900958467, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 17460 + }, + { + "epoch": 13.946485623003195, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 17461 + }, + { + "epoch": 13.947284345047922, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 17462 + }, + { + "epoch": 13.948083067092652, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 17463 + }, + { + "epoch": 13.94888178913738, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 17464 + }, + { + "epoch": 13.949680511182109, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 17465 + }, + { + "epoch": 13.950479233226837, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 17466 + }, + { + "epoch": 13.951277955271566, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 17467 + }, + { + "epoch": 13.952076677316294, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 17468 + }, + { + "epoch": 13.952875399361023, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 17469 + }, + { + "epoch": 13.95367412140575, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 17470 + }, + { + "epoch": 13.954472843450478, + "grad_norm": 0.095703125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 17471 + }, + { + "epoch": 13.955271565495208, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 17472 + }, + { + "epoch": 13.956070287539935, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 17473 + }, + { + "epoch": 13.956869009584665, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 17474 + }, + { + "epoch": 13.957667731629392, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 17475 + }, + { + "epoch": 13.958466453674122, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17476 + }, + { + "epoch": 13.95926517571885, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 17477 + }, + { + "epoch": 13.960063897763579, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 17478 + }, + { + "epoch": 13.960862619808307, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 17479 + }, + { + "epoch": 13.961661341853034, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 17480 + }, + { + "epoch": 13.962460063897764, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 17481 + }, + { + "epoch": 13.963258785942491, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17482 + }, + { + "epoch": 13.96405750798722, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 17483 + }, + { + "epoch": 13.964856230031948, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 17484 + }, + { + "epoch": 13.965654952076678, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17485 + }, + { + "epoch": 13.966453674121405, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 17486 + }, + { + "epoch": 13.967252396166135, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 17487 + }, + { + "epoch": 13.968051118210862, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 17488 + }, + { + "epoch": 13.968849840255592, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 17489 + }, + { + "epoch": 13.96964856230032, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 17490 + }, + { + "epoch": 13.970447284345047, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 17491 + }, + { + "epoch": 13.971246006389777, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 17492 + }, + { + "epoch": 13.972044728434504, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 17493 + }, + { + "epoch": 13.972843450479234, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17494 + }, + { + "epoch": 13.973642172523961, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 17495 + }, + { + "epoch": 13.97444089456869, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 17496 + }, + { + "epoch": 13.975239616613418, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 17497 + }, + { + "epoch": 13.976038338658148, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 17498 + }, + { + "epoch": 13.976837060702875, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 17499 + }, + { + "epoch": 13.977635782747605, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 17500 + }, + { + "epoch": 13.978434504792332, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 17501 + }, + { + "epoch": 13.97923322683706, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 17502 + }, + { + "epoch": 13.98003194888179, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 17503 + }, + { + "epoch": 13.980830670926517, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17504 + }, + { + "epoch": 13.981629392971247, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 17505 + }, + { + "epoch": 13.982428115015974, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 17506 + }, + { + "epoch": 13.983226837060704, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 17507 + }, + { + "epoch": 13.984025559105431, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 17508 + }, + { + "epoch": 13.98482428115016, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 17509 + }, + { + "epoch": 13.985623003194888, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 17510 + }, + { + "epoch": 13.986421725239616, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 17511 + }, + { + "epoch": 13.987220447284345, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 17512 + }, + { + "epoch": 13.988019169329073, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 17513 + }, + { + "epoch": 13.988817891373802, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 17514 + }, + { + "epoch": 13.98961661341853, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 17515 + }, + { + "epoch": 13.99041533546326, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 17516 + }, + { + "epoch": 13.991214057507987, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 17517 + }, + { + "epoch": 13.992012779552716, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 17518 + }, + { + "epoch": 13.992811501597444, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 17519 + }, + { + "epoch": 13.993610223642172, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17520 + }, + { + "epoch": 13.994408945686901, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 17521 + }, + { + "epoch": 13.995207667731629, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 17522 + }, + { + "epoch": 13.996006389776358, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 17523 + }, + { + "epoch": 13.996805111821086, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 17524 + }, + { + "epoch": 13.997603833865815, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 17525 + }, + { + "epoch": 13.998402555910543, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17526 + }, + { + "epoch": 13.999201277955272, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 17527 + }, + { + "epoch": 14.0, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 17528 + }, + { + "epoch": 14.000798722044728, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 17529 + }, + { + "epoch": 14.001597444089457, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 17530 + }, + { + "epoch": 14.002396166134185, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 17531 + }, + { + "epoch": 14.003194888178914, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 17532 + }, + { + "epoch": 14.003993610223642, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 17533 + }, + { + "epoch": 14.004792332268371, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 17534 + }, + { + "epoch": 14.005591054313099, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 17535 + }, + { + "epoch": 14.006389776357828, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 17536 + }, + { + "epoch": 14.007188498402556, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17537 + }, + { + "epoch": 14.007987220447284, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 17538 + }, + { + "epoch": 14.008785942492013, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 17539 + }, + { + "epoch": 14.00958466453674, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 17540 + }, + { + "epoch": 14.01038338658147, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 17541 + }, + { + "epoch": 14.011182108626198, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 17542 + }, + { + "epoch": 14.011980830670927, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 17543 + }, + { + "epoch": 14.012779552715655, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 17544 + }, + { + "epoch": 14.013578274760384, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 17545 + }, + { + "epoch": 14.014376996805112, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 17546 + }, + { + "epoch": 14.01517571884984, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 17547 + }, + { + "epoch": 14.015974440894569, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 17548 + }, + { + "epoch": 14.016773162939296, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 17549 + }, + { + "epoch": 14.017571884984026, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 17550 + }, + { + "epoch": 14.018370607028753, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 17551 + }, + { + "epoch": 14.019169329073483, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 17552 + }, + { + "epoch": 14.01996805111821, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17553 + }, + { + "epoch": 14.02076677316294, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17554 + }, + { + "epoch": 14.021565495207668, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 17555 + }, + { + "epoch": 14.022364217252397, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 17556 + }, + { + "epoch": 14.023162939297125, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 17557 + }, + { + "epoch": 14.023961661341852, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 17558 + }, + { + "epoch": 14.024760383386582, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 17559 + }, + { + "epoch": 14.02555910543131, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 17560 + }, + { + "epoch": 14.026357827476039, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 17561 + }, + { + "epoch": 14.027156549520766, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 17562 + }, + { + "epoch": 14.027955271565496, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 17563 + }, + { + "epoch": 14.028753993610223, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 17564 + }, + { + "epoch": 14.029552715654953, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 17565 + }, + { + "epoch": 14.03035143769968, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 17566 + }, + { + "epoch": 14.031150159744408, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 17567 + }, + { + "epoch": 14.031948881789138, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 17568 + }, + { + "epoch": 14.032747603833865, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 17569 + }, + { + "epoch": 14.033546325878595, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 17570 + }, + { + "epoch": 14.034345047923322, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 17571 + }, + { + "epoch": 14.035143769968052, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 17572 + }, + { + "epoch": 14.03594249201278, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 17573 + }, + { + "epoch": 14.036741214057509, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 17574 + }, + { + "epoch": 14.037539936102236, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 17575 + }, + { + "epoch": 14.038338658146964, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 17576 + }, + { + "epoch": 14.039137380191693, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 17577 + }, + { + "epoch": 14.039936102236421, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 17578 + }, + { + "epoch": 14.04073482428115, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 17579 + }, + { + "epoch": 14.041533546325878, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 17580 + }, + { + "epoch": 14.042332268370608, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 17581 + }, + { + "epoch": 14.043130990415335, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 17582 + }, + { + "epoch": 14.043929712460065, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 17583 + }, + { + "epoch": 14.044728434504792, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17584 + }, + { + "epoch": 14.04552715654952, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 17585 + }, + { + "epoch": 14.04632587859425, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 17586 + }, + { + "epoch": 14.047124600638977, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 17587 + }, + { + "epoch": 14.047923322683706, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 17588 + }, + { + "epoch": 14.048722044728434, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 17589 + }, + { + "epoch": 14.049520766773163, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17590 + }, + { + "epoch": 14.050319488817891, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17591 + }, + { + "epoch": 14.05111821086262, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 17592 + }, + { + "epoch": 14.051916932907348, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 17593 + }, + { + "epoch": 14.052715654952078, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 17594 + }, + { + "epoch": 14.053514376996805, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17595 + }, + { + "epoch": 14.054313099041533, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 17596 + }, + { + "epoch": 14.055111821086262, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 17597 + }, + { + "epoch": 14.05591054313099, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 17598 + }, + { + "epoch": 14.05670926517572, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 17599 + }, + { + "epoch": 14.057507987220447, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 17600 + }, + { + "epoch": 14.058306709265176, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17601 + }, + { + "epoch": 14.059105431309904, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 17602 + }, + { + "epoch": 14.059904153354633, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 17603 + }, + { + "epoch": 14.060702875399361, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 17604 + }, + { + "epoch": 14.061501597444089, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 17605 + }, + { + "epoch": 14.062300319488818, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 17606 + }, + { + "epoch": 14.063099041533546, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17607 + }, + { + "epoch": 14.063897763578275, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 17608 + }, + { + "epoch": 14.064696485623003, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 17609 + }, + { + "epoch": 14.065495207667732, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 17610 + }, + { + "epoch": 14.06629392971246, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 17611 + }, + { + "epoch": 14.06709265175719, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 17612 + }, + { + "epoch": 14.067891373801917, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 17613 + }, + { + "epoch": 14.068690095846645, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17614 + }, + { + "epoch": 14.069488817891374, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 17615 + }, + { + "epoch": 14.070287539936102, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 17616 + }, + { + "epoch": 14.071086261980831, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 17617 + }, + { + "epoch": 14.071884984025559, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 17618 + }, + { + "epoch": 14.072683706070288, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 17619 + }, + { + "epoch": 14.073482428115016, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 17620 + }, + { + "epoch": 14.074281150159745, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 17621 + }, + { + "epoch": 14.075079872204473, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 17622 + }, + { + "epoch": 14.0758785942492, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 17623 + }, + { + "epoch": 14.07667731629393, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 17624 + }, + { + "epoch": 14.077476038338657, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17625 + }, + { + "epoch": 14.078274760383387, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 17626 + }, + { + "epoch": 14.079073482428115, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 17627 + }, + { + "epoch": 14.079872204472844, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 17628 + }, + { + "epoch": 14.080670926517572, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 17629 + }, + { + "epoch": 14.081469648562301, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17630 + }, + { + "epoch": 14.082268370607029, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 17631 + }, + { + "epoch": 14.083067092651758, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 17632 + }, + { + "epoch": 14.083865814696486, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 17633 + }, + { + "epoch": 14.084664536741213, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 17634 + }, + { + "epoch": 14.085463258785943, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 17635 + }, + { + "epoch": 14.08626198083067, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 17636 + }, + { + "epoch": 14.0870607028754, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 17637 + }, + { + "epoch": 14.087859424920127, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 17638 + }, + { + "epoch": 14.088658146964857, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 17639 + }, + { + "epoch": 14.089456869009584, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 17640 + }, + { + "epoch": 14.090255591054314, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 17641 + }, + { + "epoch": 14.091054313099042, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 17642 + }, + { + "epoch": 14.09185303514377, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 17643 + }, + { + "epoch": 14.092651757188499, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 17644 + }, + { + "epoch": 14.093450479233226, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 17645 + }, + { + "epoch": 14.094249201277956, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 17646 + }, + { + "epoch": 14.095047923322683, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 17647 + }, + { + "epoch": 14.095846645367413, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 17648 + }, + { + "epoch": 14.09664536741214, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 17649 + }, + { + "epoch": 14.09744408945687, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 17650 + }, + { + "epoch": 14.098242811501597, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 17651 + }, + { + "epoch": 14.099041533546325, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 17652 + }, + { + "epoch": 14.099840255591054, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 17653 + }, + { + "epoch": 14.100638977635782, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 17654 + }, + { + "epoch": 14.101437699680512, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 17655 + }, + { + "epoch": 14.10223642172524, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 17656 + }, + { + "epoch": 14.103035143769969, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 17657 + }, + { + "epoch": 14.103833865814696, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 17658 + }, + { + "epoch": 14.104632587859426, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 17659 + }, + { + "epoch": 14.105431309904153, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 17660 + }, + { + "epoch": 14.106230031948881, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17661 + }, + { + "epoch": 14.10702875399361, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 17662 + }, + { + "epoch": 14.107827476038338, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 17663 + }, + { + "epoch": 14.108626198083067, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 17664 + }, + { + "epoch": 14.109424920127795, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 17665 + }, + { + "epoch": 14.110223642172524, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17666 + }, + { + "epoch": 14.111022364217252, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 17667 + }, + { + "epoch": 14.111821086261982, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 17668 + }, + { + "epoch": 14.11261980830671, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 17669 + }, + { + "epoch": 14.113418530351439, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 17670 + }, + { + "epoch": 14.114217252396166, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 17671 + }, + { + "epoch": 14.115015974440894, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 17672 + }, + { + "epoch": 14.115814696485623, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17673 + }, + { + "epoch": 14.116613418530351, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 17674 + }, + { + "epoch": 14.11741214057508, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 17675 + }, + { + "epoch": 14.118210862619808, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 17676 + }, + { + "epoch": 14.119009584664537, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 17677 + }, + { + "epoch": 14.119808306709265, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 17678 + }, + { + "epoch": 14.120607028753994, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 17679 + }, + { + "epoch": 14.121405750798722, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 17680 + }, + { + "epoch": 14.12220447284345, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 17681 + }, + { + "epoch": 14.12300319488818, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 17682 + }, + { + "epoch": 14.123801916932907, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 17683 + }, + { + "epoch": 14.124600638977636, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 17684 + }, + { + "epoch": 14.125399361022364, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 17685 + }, + { + "epoch": 14.126198083067093, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 17686 + }, + { + "epoch": 14.12699680511182, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 17687 + }, + { + "epoch": 14.12779552715655, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 17688 + }, + { + "epoch": 14.128594249201278, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 17689 + }, + { + "epoch": 14.129392971246006, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 17690 + }, + { + "epoch": 14.130191693290735, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 17691 + }, + { + "epoch": 14.130990415335463, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 17692 + }, + { + "epoch": 14.131789137380192, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 17693 + }, + { + "epoch": 14.13258785942492, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17694 + }, + { + "epoch": 14.133386581469649, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 17695 + }, + { + "epoch": 14.134185303514377, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17696 + }, + { + "epoch": 14.134984025559106, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17697 + }, + { + "epoch": 14.135782747603834, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 17698 + }, + { + "epoch": 14.136581469648561, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 17699 + }, + { + "epoch": 14.13738019169329, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 17700 + }, + { + "epoch": 14.138178913738018, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 17701 + }, + { + "epoch": 14.138977635782748, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 17702 + }, + { + "epoch": 14.139776357827476, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 17703 + }, + { + "epoch": 14.140575079872205, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 17704 + }, + { + "epoch": 14.141373801916933, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 17705 + }, + { + "epoch": 14.142172523961662, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 17706 + }, + { + "epoch": 14.14297124600639, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 17707 + }, + { + "epoch": 14.143769968051119, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 17708 + }, + { + "epoch": 14.144568690095847, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 17709 + }, + { + "epoch": 14.145367412140574, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 17710 + }, + { + "epoch": 14.146166134185304, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 17711 + }, + { + "epoch": 14.146964856230031, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 17712 + }, + { + "epoch": 14.14776357827476, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 17713 + }, + { + "epoch": 14.148562300319488, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 17714 + }, + { + "epoch": 14.149361022364218, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 17715 + }, + { + "epoch": 14.150159744408946, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 17716 + }, + { + "epoch": 14.150958466453675, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17717 + }, + { + "epoch": 14.151757188498403, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 17718 + }, + { + "epoch": 14.15255591054313, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 17719 + }, + { + "epoch": 14.15335463258786, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 17720 + }, + { + "epoch": 14.154153354632587, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 17721 + }, + { + "epoch": 14.154952076677317, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 17722 + }, + { + "epoch": 14.155750798722044, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 17723 + }, + { + "epoch": 14.156549520766774, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 17724 + }, + { + "epoch": 14.157348242811501, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 17725 + }, + { + "epoch": 14.15814696485623, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 17726 + }, + { + "epoch": 14.158945686900958, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 17727 + }, + { + "epoch": 14.159744408945686, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 17728 + }, + { + "epoch": 14.160543130990416, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 17729 + }, + { + "epoch": 14.161341853035143, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 17730 + }, + { + "epoch": 14.162140575079873, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 17731 + }, + { + "epoch": 14.1629392971246, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 17732 + }, + { + "epoch": 14.16373801916933, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 17733 + }, + { + "epoch": 14.164536741214057, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17734 + }, + { + "epoch": 14.165335463258787, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 17735 + }, + { + "epoch": 14.166134185303514, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17736 + }, + { + "epoch": 14.166932907348242, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 17737 + }, + { + "epoch": 14.167731629392971, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 17738 + }, + { + "epoch": 14.168530351437699, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 17739 + }, + { + "epoch": 14.169329073482428, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 17740 + }, + { + "epoch": 14.170127795527156, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 17741 + }, + { + "epoch": 14.170926517571885, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17742 + }, + { + "epoch": 14.171725239616613, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17743 + }, + { + "epoch": 14.172523961661343, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 17744 + }, + { + "epoch": 14.17332268370607, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 17745 + }, + { + "epoch": 14.1741214057508, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 17746 + }, + { + "epoch": 14.174920127795527, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 17747 + }, + { + "epoch": 14.175718849840255, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 17748 + }, + { + "epoch": 14.176517571884984, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 17749 + }, + { + "epoch": 14.177316293929712, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17750 + }, + { + "epoch": 14.178115015974441, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 17751 + }, + { + "epoch": 14.178913738019169, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 17752 + }, + { + "epoch": 14.179712460063898, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 17753 + }, + { + "epoch": 14.180511182108626, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 17754 + }, + { + "epoch": 14.181309904153355, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 17755 + }, + { + "epoch": 14.182108626198083, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17756 + }, + { + "epoch": 14.18290734824281, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 17757 + }, + { + "epoch": 14.18370607028754, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 17758 + }, + { + "epoch": 14.184504792332268, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17759 + }, + { + "epoch": 14.185303514376997, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 17760 + }, + { + "epoch": 14.186102236421725, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 17761 + }, + { + "epoch": 14.186900958466454, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 17762 + }, + { + "epoch": 14.187699680511182, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 17763 + }, + { + "epoch": 14.188498402555911, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17764 + }, + { + "epoch": 14.189297124600639, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 17765 + }, + { + "epoch": 14.190095846645367, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 17766 + }, + { + "epoch": 14.190894568690096, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 17767 + }, + { + "epoch": 14.191693290734824, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 17768 + }, + { + "epoch": 14.192492012779553, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 17769 + }, + { + "epoch": 14.19329073482428, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 17770 + }, + { + "epoch": 14.19408945686901, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 17771 + }, + { + "epoch": 14.194888178913738, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 17772 + }, + { + "epoch": 14.195686900958467, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 17773 + }, + { + "epoch": 14.196485623003195, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 17774 + }, + { + "epoch": 14.197284345047922, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 17775 + }, + { + "epoch": 14.198083067092652, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 17776 + }, + { + "epoch": 14.19888178913738, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 17777 + }, + { + "epoch": 14.199680511182109, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 17778 + }, + { + "epoch": 14.200479233226837, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 17779 + }, + { + "epoch": 14.201277955271566, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 17780 + }, + { + "epoch": 14.202076677316294, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 17781 + }, + { + "epoch": 14.202875399361023, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17782 + }, + { + "epoch": 14.20367412140575, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 17783 + }, + { + "epoch": 14.204472843450478, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 17784 + }, + { + "epoch": 14.205271565495208, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 17785 + }, + { + "epoch": 14.206070287539935, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 17786 + }, + { + "epoch": 14.206869009584665, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 17787 + }, + { + "epoch": 14.207667731629392, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 17788 + }, + { + "epoch": 14.208466453674122, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 17789 + }, + { + "epoch": 14.20926517571885, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 17790 + }, + { + "epoch": 14.210063897763579, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 17791 + }, + { + "epoch": 14.210862619808307, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 17792 + }, + { + "epoch": 14.211661341853036, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 17793 + }, + { + "epoch": 14.212460063897764, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 17794 + }, + { + "epoch": 14.213258785942491, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 17795 + }, + { + "epoch": 14.21405750798722, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17796 + }, + { + "epoch": 14.214856230031948, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 17797 + }, + { + "epoch": 14.215654952076678, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 17798 + }, + { + "epoch": 14.216453674121405, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 17799 + }, + { + "epoch": 14.217252396166135, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 17800 + }, + { + "epoch": 14.218051118210862, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 17801 + }, + { + "epoch": 14.218849840255592, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 17802 + }, + { + "epoch": 14.21964856230032, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 17803 + }, + { + "epoch": 14.220447284345047, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 17804 + }, + { + "epoch": 14.221246006389777, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 17805 + }, + { + "epoch": 14.222044728434504, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 17806 + }, + { + "epoch": 14.222843450479234, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 17807 + }, + { + "epoch": 14.223642172523961, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 17808 + }, + { + "epoch": 14.22444089456869, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17809 + }, + { + "epoch": 14.225239616613418, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17810 + }, + { + "epoch": 14.226038338658148, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 17811 + }, + { + "epoch": 14.226837060702875, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 17812 + }, + { + "epoch": 14.227635782747603, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 17813 + }, + { + "epoch": 14.228434504792332, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 17814 + }, + { + "epoch": 14.22923322683706, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 17815 + }, + { + "epoch": 14.23003194888179, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 17816 + }, + { + "epoch": 14.230830670926517, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 17817 + }, + { + "epoch": 14.231629392971247, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 17818 + }, + { + "epoch": 14.232428115015974, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 17819 + }, + { + "epoch": 14.233226837060704, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 17820 + }, + { + "epoch": 14.234025559105431, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 17821 + }, + { + "epoch": 14.23482428115016, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 17822 + }, + { + "epoch": 14.235623003194888, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17823 + }, + { + "epoch": 14.236421725239616, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 17824 + }, + { + "epoch": 14.237220447284345, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 17825 + }, + { + "epoch": 14.238019169329073, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 17826 + }, + { + "epoch": 14.238817891373802, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 17827 + }, + { + "epoch": 14.23961661341853, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 17828 + }, + { + "epoch": 14.24041533546326, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 17829 + }, + { + "epoch": 14.241214057507987, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 17830 + }, + { + "epoch": 14.242012779552716, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 17831 + }, + { + "epoch": 14.242811501597444, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 17832 + }, + { + "epoch": 14.243610223642172, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 17833 + }, + { + "epoch": 14.244408945686901, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 17834 + }, + { + "epoch": 14.245207667731629, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 17835 + }, + { + "epoch": 14.246006389776358, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 17836 + }, + { + "epoch": 14.246805111821086, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 17837 + }, + { + "epoch": 14.247603833865815, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 17838 + }, + { + "epoch": 14.248402555910543, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 17839 + }, + { + "epoch": 14.249201277955272, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 17840 + }, + { + "epoch": 14.25, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 17841 + }, + { + "epoch": 14.250798722044728, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 17842 + }, + { + "epoch": 14.251597444089457, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 17843 + }, + { + "epoch": 14.252396166134185, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 17844 + }, + { + "epoch": 14.253194888178914, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 17845 + }, + { + "epoch": 14.253993610223642, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 17846 + }, + { + "epoch": 14.254792332268371, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 17847 + }, + { + "epoch": 14.255591054313099, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 17848 + }, + { + "epoch": 14.256389776357828, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 17849 + }, + { + "epoch": 14.257188498402556, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 17850 + }, + { + "epoch": 14.257987220447284, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 17851 + }, + { + "epoch": 14.258785942492013, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 17852 + }, + { + "epoch": 14.25958466453674, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 17853 + }, + { + "epoch": 14.26038338658147, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 17854 + }, + { + "epoch": 14.261182108626198, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 17855 + }, + { + "epoch": 14.261980830670927, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 17856 + }, + { + "epoch": 14.262779552715655, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 17857 + }, + { + "epoch": 14.263578274760384, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 17858 + }, + { + "epoch": 14.264376996805112, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 17859 + }, + { + "epoch": 14.26517571884984, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 17860 + }, + { + "epoch": 14.265974440894569, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 17861 + }, + { + "epoch": 14.266773162939296, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 17862 + }, + { + "epoch": 14.267571884984026, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 17863 + }, + { + "epoch": 14.268370607028753, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 17864 + }, + { + "epoch": 14.269169329073483, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 17865 + }, + { + "epoch": 14.26996805111821, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 17866 + }, + { + "epoch": 14.27076677316294, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 17867 + }, + { + "epoch": 14.271565495207668, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 17868 + }, + { + "epoch": 14.272364217252397, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 17869 + }, + { + "epoch": 14.273162939297125, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 17870 + }, + { + "epoch": 14.273961661341852, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 17871 + }, + { + "epoch": 14.274760383386582, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 17872 + }, + { + "epoch": 14.27555910543131, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 17873 + }, + { + "epoch": 14.276357827476039, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 17874 + }, + { + "epoch": 14.277156549520766, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 17875 + }, + { + "epoch": 14.277955271565496, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17876 + }, + { + "epoch": 14.278753993610223, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 17877 + }, + { + "epoch": 14.279552715654953, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 17878 + }, + { + "epoch": 14.28035143769968, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 17879 + }, + { + "epoch": 14.281150159744408, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 17880 + }, + { + "epoch": 14.281948881789138, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17881 + }, + { + "epoch": 14.282747603833865, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 17882 + }, + { + "epoch": 14.283546325878595, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 17883 + }, + { + "epoch": 14.284345047923322, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 17884 + }, + { + "epoch": 14.285143769968052, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 17885 + }, + { + "epoch": 14.28594249201278, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 17886 + }, + { + "epoch": 14.286741214057509, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 17887 + }, + { + "epoch": 14.287539936102236, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 17888 + }, + { + "epoch": 14.288338658146964, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 17889 + }, + { + "epoch": 14.289137380191693, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 17890 + }, + { + "epoch": 14.289936102236421, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 17891 + }, + { + "epoch": 14.29073482428115, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 17892 + }, + { + "epoch": 14.291533546325878, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 17893 + }, + { + "epoch": 14.292332268370608, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 17894 + }, + { + "epoch": 14.293130990415335, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 17895 + }, + { + "epoch": 14.293929712460065, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 17896 + }, + { + "epoch": 14.294728434504792, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 17897 + }, + { + "epoch": 14.295527156549522, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 17898 + }, + { + "epoch": 14.29632587859425, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 17899 + }, + { + "epoch": 14.297124600638977, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 17900 + }, + { + "epoch": 14.297923322683706, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 17901 + }, + { + "epoch": 14.298722044728434, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 17902 + }, + { + "epoch": 14.299520766773163, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 17903 + }, + { + "epoch": 14.300319488817891, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 17904 + }, + { + "epoch": 14.30111821086262, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 17905 + }, + { + "epoch": 14.301916932907348, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 17906 + }, + { + "epoch": 14.302715654952078, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 17907 + }, + { + "epoch": 14.303514376996805, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 17908 + }, + { + "epoch": 14.304313099041533, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17909 + }, + { + "epoch": 14.305111821086262, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 17910 + }, + { + "epoch": 14.30591054313099, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 17911 + }, + { + "epoch": 14.30670926517572, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 17912 + }, + { + "epoch": 14.307507987220447, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 17913 + }, + { + "epoch": 14.308306709265176, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 17914 + }, + { + "epoch": 14.309105431309904, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 17915 + }, + { + "epoch": 14.309904153354633, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 17916 + }, + { + "epoch": 14.310702875399361, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 17917 + }, + { + "epoch": 14.311501597444089, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 17918 + }, + { + "epoch": 14.312300319488818, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 17919 + }, + { + "epoch": 14.313099041533546, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 17920 + }, + { + "epoch": 14.313897763578275, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 17921 + }, + { + "epoch": 14.314696485623003, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 17922 + }, + { + "epoch": 14.315495207667732, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 17923 + }, + { + "epoch": 14.31629392971246, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 17924 + }, + { + "epoch": 14.31709265175719, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 17925 + }, + { + "epoch": 14.317891373801917, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17926 + }, + { + "epoch": 14.318690095846645, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 17927 + }, + { + "epoch": 14.319488817891374, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 17928 + }, + { + "epoch": 14.320287539936102, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 17929 + }, + { + "epoch": 14.321086261980831, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 17930 + }, + { + "epoch": 14.321884984025559, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 17931 + }, + { + "epoch": 14.322683706070288, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 17932 + }, + { + "epoch": 14.323482428115016, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 17933 + }, + { + "epoch": 14.324281150159745, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 17934 + }, + { + "epoch": 14.325079872204473, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 17935 + }, + { + "epoch": 14.3258785942492, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 17936 + }, + { + "epoch": 14.32667731629393, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 17937 + }, + { + "epoch": 14.327476038338657, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 17938 + }, + { + "epoch": 14.328274760383387, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 17939 + }, + { + "epoch": 14.329073482428115, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 17940 + }, + { + "epoch": 14.329872204472844, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 17941 + }, + { + "epoch": 14.330670926517572, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 17942 + }, + { + "epoch": 14.331469648562301, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 17943 + }, + { + "epoch": 14.332268370607029, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 17944 + }, + { + "epoch": 14.333067092651758, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 17945 + }, + { + "epoch": 14.333865814696486, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 17946 + }, + { + "epoch": 14.334664536741213, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 17947 + }, + { + "epoch": 14.335463258785943, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 17948 + }, + { + "epoch": 14.33626198083067, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 17949 + }, + { + "epoch": 14.3370607028754, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 17950 + }, + { + "epoch": 14.337859424920127, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 17951 + }, + { + "epoch": 14.338658146964857, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17952 + }, + { + "epoch": 14.339456869009584, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 17953 + }, + { + "epoch": 14.340255591054314, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 17954 + }, + { + "epoch": 14.341054313099042, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 17955 + }, + { + "epoch": 14.34185303514377, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 17956 + }, + { + "epoch": 14.342651757188499, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 17957 + }, + { + "epoch": 14.343450479233226, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 17958 + }, + { + "epoch": 14.344249201277956, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 17959 + }, + { + "epoch": 14.345047923322683, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 17960 + }, + { + "epoch": 14.345846645367413, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 17961 + }, + { + "epoch": 14.34664536741214, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 17962 + }, + { + "epoch": 14.34744408945687, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 17963 + }, + { + "epoch": 14.348242811501597, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 17964 + }, + { + "epoch": 14.349041533546325, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 17965 + }, + { + "epoch": 14.349840255591054, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 17966 + }, + { + "epoch": 14.350638977635782, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 17967 + }, + { + "epoch": 14.351437699680512, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 17968 + }, + { + "epoch": 14.35223642172524, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 17969 + }, + { + "epoch": 14.353035143769969, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 17970 + }, + { + "epoch": 14.353833865814696, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 17971 + }, + { + "epoch": 14.354632587859426, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 17972 + }, + { + "epoch": 14.355431309904153, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 17973 + }, + { + "epoch": 14.356230031948883, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 17974 + }, + { + "epoch": 14.35702875399361, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 17975 + }, + { + "epoch": 14.357827476038338, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 17976 + }, + { + "epoch": 14.358626198083067, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 17977 + }, + { + "epoch": 14.359424920127795, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 17978 + }, + { + "epoch": 14.360223642172524, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 17979 + }, + { + "epoch": 14.361022364217252, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 17980 + }, + { + "epoch": 14.361821086261982, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 17981 + }, + { + "epoch": 14.36261980830671, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 17982 + }, + { + "epoch": 14.363418530351439, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 17983 + }, + { + "epoch": 14.364217252396166, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 17984 + }, + { + "epoch": 14.365015974440894, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 17985 + }, + { + "epoch": 14.365814696485623, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 17986 + }, + { + "epoch": 14.366613418530351, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 17987 + }, + { + "epoch": 14.36741214057508, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 17988 + }, + { + "epoch": 14.368210862619808, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 17989 + }, + { + "epoch": 14.369009584664537, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 17990 + }, + { + "epoch": 14.369808306709265, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 17991 + }, + { + "epoch": 14.370607028753994, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 17992 + }, + { + "epoch": 14.371405750798722, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 17993 + }, + { + "epoch": 14.37220447284345, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 17994 + }, + { + "epoch": 14.37300319488818, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 17995 + }, + { + "epoch": 14.373801916932907, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 17996 + }, + { + "epoch": 14.374600638977636, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 17997 + }, + { + "epoch": 14.375399361022364, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 17998 + }, + { + "epoch": 14.376198083067093, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 17999 + }, + { + "epoch": 14.37699680511182, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 18000 + }, + { + "epoch": 14.37779552715655, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 18001 + }, + { + "epoch": 14.378594249201278, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 18002 + }, + { + "epoch": 14.379392971246006, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 18003 + }, + { + "epoch": 14.380191693290735, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 18004 + }, + { + "epoch": 14.380990415335463, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 18005 + }, + { + "epoch": 14.381789137380192, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 18006 + }, + { + "epoch": 14.38258785942492, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 18007 + }, + { + "epoch": 14.383386581469649, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 18008 + }, + { + "epoch": 14.384185303514377, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 18009 + }, + { + "epoch": 14.384984025559106, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 18010 + }, + { + "epoch": 14.385782747603834, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 18011 + }, + { + "epoch": 14.386581469648561, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 18012 + }, + { + "epoch": 14.38738019169329, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 18013 + }, + { + "epoch": 14.388178913738018, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 18014 + }, + { + "epoch": 14.388977635782748, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 18015 + }, + { + "epoch": 14.389776357827476, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 18016 + }, + { + "epoch": 14.390575079872205, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 18017 + }, + { + "epoch": 14.391373801916933, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 18018 + }, + { + "epoch": 14.392172523961662, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 18019 + }, + { + "epoch": 14.39297124600639, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 18020 + }, + { + "epoch": 14.393769968051117, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 18021 + }, + { + "epoch": 14.394568690095847, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 18022 + }, + { + "epoch": 14.395367412140574, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 18023 + }, + { + "epoch": 14.396166134185304, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 18024 + }, + { + "epoch": 14.396964856230031, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18025 + }, + { + "epoch": 14.39776357827476, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18026 + }, + { + "epoch": 14.398562300319488, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 18027 + }, + { + "epoch": 14.399361022364218, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 18028 + }, + { + "epoch": 14.400159744408946, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 18029 + }, + { + "epoch": 14.400958466453675, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 18030 + }, + { + "epoch": 14.401757188498403, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18031 + }, + { + "epoch": 14.40255591054313, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 18032 + }, + { + "epoch": 14.40335463258786, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 18033 + }, + { + "epoch": 14.404153354632587, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 18034 + }, + { + "epoch": 14.404952076677317, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 18035 + }, + { + "epoch": 14.405750798722044, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 18036 + }, + { + "epoch": 14.406549520766774, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 18037 + }, + { + "epoch": 14.407348242811501, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 18038 + }, + { + "epoch": 14.40814696485623, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 18039 + }, + { + "epoch": 14.408945686900958, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 18040 + }, + { + "epoch": 14.409744408945686, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 18041 + }, + { + "epoch": 14.410543130990416, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 18042 + }, + { + "epoch": 14.411341853035143, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 18043 + }, + { + "epoch": 14.412140575079873, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18044 + }, + { + "epoch": 14.4129392971246, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 18045 + }, + { + "epoch": 14.41373801916933, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 18046 + }, + { + "epoch": 14.414536741214057, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 18047 + }, + { + "epoch": 14.415335463258787, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 18048 + }, + { + "epoch": 14.416134185303514, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 18049 + }, + { + "epoch": 14.416932907348242, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 18050 + }, + { + "epoch": 14.417731629392971, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 18051 + }, + { + "epoch": 14.418530351437699, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 18052 + }, + { + "epoch": 14.419329073482428, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 18053 + }, + { + "epoch": 14.420127795527156, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 18054 + }, + { + "epoch": 14.420926517571885, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 18055 + }, + { + "epoch": 14.421725239616613, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 18056 + }, + { + "epoch": 14.422523961661343, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 18057 + }, + { + "epoch": 14.42332268370607, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18058 + }, + { + "epoch": 14.4241214057508, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 18059 + }, + { + "epoch": 14.424920127795527, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 18060 + }, + { + "epoch": 14.425718849840255, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 18061 + }, + { + "epoch": 14.426517571884984, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 18062 + }, + { + "epoch": 14.427316293929712, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 18063 + }, + { + "epoch": 14.428115015974441, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 18064 + }, + { + "epoch": 14.428913738019169, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 18065 + }, + { + "epoch": 14.429712460063898, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 18066 + }, + { + "epoch": 14.430511182108626, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 18067 + }, + { + "epoch": 14.431309904153355, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 18068 + }, + { + "epoch": 14.432108626198083, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 18069 + }, + { + "epoch": 14.43290734824281, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 18070 + }, + { + "epoch": 14.43370607028754, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 18071 + }, + { + "epoch": 14.434504792332268, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 18072 + }, + { + "epoch": 14.435303514376997, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 18073 + }, + { + "epoch": 14.436102236421725, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 18074 + }, + { + "epoch": 14.436900958466454, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 18075 + }, + { + "epoch": 14.437699680511182, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 18076 + }, + { + "epoch": 14.438498402555911, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 18077 + }, + { + "epoch": 14.439297124600639, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 18078 + }, + { + "epoch": 14.440095846645367, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 18079 + }, + { + "epoch": 14.440894568690096, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 18080 + }, + { + "epoch": 14.441693290734824, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 18081 + }, + { + "epoch": 14.442492012779553, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 18082 + }, + { + "epoch": 14.44329073482428, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 18083 + }, + { + "epoch": 14.44408945686901, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 18084 + }, + { + "epoch": 14.444888178913738, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 18085 + }, + { + "epoch": 14.445686900958467, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 18086 + }, + { + "epoch": 14.446485623003195, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 18087 + }, + { + "epoch": 14.447284345047922, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 18088 + }, + { + "epoch": 14.448083067092652, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 18089 + }, + { + "epoch": 14.44888178913738, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 18090 + }, + { + "epoch": 14.449680511182109, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 18091 + }, + { + "epoch": 14.450479233226837, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 18092 + }, + { + "epoch": 14.451277955271566, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 18093 + }, + { + "epoch": 14.452076677316294, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18094 + }, + { + "epoch": 14.452875399361023, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 18095 + }, + { + "epoch": 14.45367412140575, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 18096 + }, + { + "epoch": 14.454472843450478, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 18097 + }, + { + "epoch": 14.455271565495208, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 18098 + }, + { + "epoch": 14.456070287539935, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 18099 + }, + { + "epoch": 14.456869009584665, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 18100 + }, + { + "epoch": 14.457667731629392, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 18101 + }, + { + "epoch": 14.458466453674122, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 18102 + }, + { + "epoch": 14.45926517571885, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 18103 + }, + { + "epoch": 14.460063897763579, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 18104 + }, + { + "epoch": 14.460862619808307, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 18105 + }, + { + "epoch": 14.461661341853036, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 18106 + }, + { + "epoch": 14.462460063897764, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 18107 + }, + { + "epoch": 14.463258785942491, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 18108 + }, + { + "epoch": 14.46405750798722, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 18109 + }, + { + "epoch": 14.464856230031948, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 18110 + }, + { + "epoch": 14.465654952076678, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 18111 + }, + { + "epoch": 14.466453674121405, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18112 + }, + { + "epoch": 14.467252396166135, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 18113 + }, + { + "epoch": 14.468051118210862, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 18114 + }, + { + "epoch": 14.468849840255592, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 18115 + }, + { + "epoch": 14.46964856230032, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 18116 + }, + { + "epoch": 14.470447284345047, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 18117 + }, + { + "epoch": 14.471246006389777, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 18118 + }, + { + "epoch": 14.472044728434504, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 18119 + }, + { + "epoch": 14.472843450479234, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 18120 + }, + { + "epoch": 14.473642172523961, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 18121 + }, + { + "epoch": 14.47444089456869, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 18122 + }, + { + "epoch": 14.475239616613418, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 18123 + }, + { + "epoch": 14.476038338658148, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 18124 + }, + { + "epoch": 14.476837060702875, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 18125 + }, + { + "epoch": 14.477635782747603, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18126 + }, + { + "epoch": 14.478434504792332, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 18127 + }, + { + "epoch": 14.47923322683706, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 18128 + }, + { + "epoch": 14.48003194888179, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 18129 + }, + { + "epoch": 14.480830670926517, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 18130 + }, + { + "epoch": 14.481629392971247, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 18131 + }, + { + "epoch": 14.482428115015974, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 18132 + }, + { + "epoch": 14.483226837060704, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 18133 + }, + { + "epoch": 14.484025559105431, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 18134 + }, + { + "epoch": 14.48482428115016, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18135 + }, + { + "epoch": 14.485623003194888, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 18136 + }, + { + "epoch": 14.486421725239616, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18137 + }, + { + "epoch": 14.487220447284345, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18138 + }, + { + "epoch": 14.488019169329073, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 18139 + }, + { + "epoch": 14.488817891373802, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 18140 + }, + { + "epoch": 14.48961661341853, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 18141 + }, + { + "epoch": 14.49041533546326, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 18142 + }, + { + "epoch": 14.491214057507987, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 18143 + }, + { + "epoch": 14.492012779552716, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 18144 + }, + { + "epoch": 14.492811501597444, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 18145 + }, + { + "epoch": 14.493610223642172, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 18146 + }, + { + "epoch": 14.494408945686901, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 18147 + }, + { + "epoch": 14.495207667731629, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 18148 + }, + { + "epoch": 14.496006389776358, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18149 + }, + { + "epoch": 14.496805111821086, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 18150 + }, + { + "epoch": 14.497603833865815, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 18151 + }, + { + "epoch": 14.498402555910543, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 18152 + }, + { + "epoch": 14.499201277955272, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 18153 + }, + { + "epoch": 14.5, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 18154 + }, + { + "epoch": 14.500798722044728, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 18155 + }, + { + "epoch": 14.501597444089457, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 18156 + }, + { + "epoch": 14.502396166134185, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 18157 + }, + { + "epoch": 14.503194888178914, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 18158 + }, + { + "epoch": 14.503993610223642, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 18159 + }, + { + "epoch": 14.504792332268371, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 18160 + }, + { + "epoch": 14.505591054313099, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 18161 + }, + { + "epoch": 14.506389776357828, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 18162 + }, + { + "epoch": 14.507188498402556, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 18163 + }, + { + "epoch": 14.507987220447284, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 18164 + }, + { + "epoch": 14.508785942492013, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 18165 + }, + { + "epoch": 14.50958466453674, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 18166 + }, + { + "epoch": 14.51038338658147, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 18167 + }, + { + "epoch": 14.511182108626198, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 18168 + }, + { + "epoch": 14.511980830670927, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 18169 + }, + { + "epoch": 14.512779552715655, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 18170 + }, + { + "epoch": 14.513578274760384, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 18171 + }, + { + "epoch": 14.514376996805112, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 18172 + }, + { + "epoch": 14.51517571884984, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 18173 + }, + { + "epoch": 14.515974440894569, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 18174 + }, + { + "epoch": 14.516773162939296, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 18175 + }, + { + "epoch": 14.517571884984026, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 18176 + }, + { + "epoch": 14.518370607028753, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 18177 + }, + { + "epoch": 14.519169329073483, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 18178 + }, + { + "epoch": 14.51996805111821, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 18179 + }, + { + "epoch": 14.52076677316294, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 18180 + }, + { + "epoch": 14.521565495207668, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18181 + }, + { + "epoch": 14.522364217252395, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 18182 + }, + { + "epoch": 14.523162939297125, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 18183 + }, + { + "epoch": 14.523961661341852, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 18184 + }, + { + "epoch": 14.524760383386582, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 18185 + }, + { + "epoch": 14.52555910543131, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 18186 + }, + { + "epoch": 14.526357827476039, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 18187 + }, + { + "epoch": 14.527156549520766, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 18188 + }, + { + "epoch": 14.527955271565496, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 18189 + }, + { + "epoch": 14.528753993610223, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 18190 + }, + { + "epoch": 14.529552715654953, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 18191 + }, + { + "epoch": 14.53035143769968, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 18192 + }, + { + "epoch": 14.531150159744408, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 18193 + }, + { + "epoch": 14.531948881789138, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 18194 + }, + { + "epoch": 14.532747603833865, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 18195 + }, + { + "epoch": 14.533546325878595, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 18196 + }, + { + "epoch": 14.534345047923322, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 18197 + }, + { + "epoch": 14.535143769968052, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 18198 + }, + { + "epoch": 14.53594249201278, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 18199 + }, + { + "epoch": 14.536741214057509, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 18200 + }, + { + "epoch": 14.537539936102236, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 18201 + }, + { + "epoch": 14.538338658146966, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 18202 + }, + { + "epoch": 14.539137380191693, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 18203 + }, + { + "epoch": 14.539936102236421, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18204 + }, + { + "epoch": 14.54073482428115, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 18205 + }, + { + "epoch": 14.541533546325878, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 18206 + }, + { + "epoch": 14.542332268370608, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 18207 + }, + { + "epoch": 14.543130990415335, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 18208 + }, + { + "epoch": 14.543929712460065, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 18209 + }, + { + "epoch": 14.544728434504792, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 18210 + }, + { + "epoch": 14.545527156549522, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 18211 + }, + { + "epoch": 14.54632587859425, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 18212 + }, + { + "epoch": 14.547124600638977, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 18213 + }, + { + "epoch": 14.547923322683706, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 18214 + }, + { + "epoch": 14.548722044728434, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 18215 + }, + { + "epoch": 14.549520766773163, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 18216 + }, + { + "epoch": 14.550319488817891, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 18217 + }, + { + "epoch": 14.55111821086262, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 18218 + }, + { + "epoch": 14.551916932907348, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 18219 + }, + { + "epoch": 14.552715654952078, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 18220 + }, + { + "epoch": 14.553514376996805, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 18221 + }, + { + "epoch": 14.554313099041533, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 18222 + }, + { + "epoch": 14.555111821086262, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 18223 + }, + { + "epoch": 14.55591054313099, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 18224 + }, + { + "epoch": 14.55670926517572, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18225 + }, + { + "epoch": 14.557507987220447, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 18226 + }, + { + "epoch": 14.558306709265176, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 18227 + }, + { + "epoch": 14.559105431309904, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1102, + "step": 18228 + }, + { + "epoch": 14.559904153354633, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 18229 + }, + { + "epoch": 14.560702875399361, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 18230 + }, + { + "epoch": 14.561501597444089, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18231 + }, + { + "epoch": 14.562300319488818, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 18232 + }, + { + "epoch": 14.563099041533546, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 18233 + }, + { + "epoch": 14.563897763578275, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 18234 + }, + { + "epoch": 14.564696485623003, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 18235 + }, + { + "epoch": 14.565495207667732, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 18236 + }, + { + "epoch": 14.56629392971246, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 18237 + }, + { + "epoch": 14.56709265175719, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18238 + }, + { + "epoch": 14.567891373801917, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 18239 + }, + { + "epoch": 14.568690095846645, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 18240 + }, + { + "epoch": 14.569488817891374, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 18241 + }, + { + "epoch": 14.570287539936102, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 18242 + }, + { + "epoch": 14.571086261980831, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 18243 + }, + { + "epoch": 14.571884984025559, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 18244 + }, + { + "epoch": 14.572683706070288, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 18245 + }, + { + "epoch": 14.573482428115016, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18246 + }, + { + "epoch": 14.574281150159745, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18247 + }, + { + "epoch": 14.575079872204473, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 18248 + }, + { + "epoch": 14.5758785942492, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 18249 + }, + { + "epoch": 14.57667731629393, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 18250 + }, + { + "epoch": 14.577476038338657, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 18251 + }, + { + "epoch": 14.578274760383387, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 18252 + }, + { + "epoch": 14.579073482428115, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 18253 + }, + { + "epoch": 14.579872204472844, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 18254 + }, + { + "epoch": 14.580670926517572, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 18255 + }, + { + "epoch": 14.581469648562301, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 18256 + }, + { + "epoch": 14.582268370607029, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 18257 + }, + { + "epoch": 14.583067092651756, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 18258 + }, + { + "epoch": 14.583865814696486, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 18259 + }, + { + "epoch": 14.584664536741213, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 18260 + }, + { + "epoch": 14.585463258785943, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 18261 + }, + { + "epoch": 14.58626198083067, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 18262 + }, + { + "epoch": 14.5870607028754, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 18263 + }, + { + "epoch": 14.587859424920127, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 18264 + }, + { + "epoch": 14.588658146964857, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 18265 + }, + { + "epoch": 14.589456869009584, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 18266 + }, + { + "epoch": 14.590255591054314, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 18267 + }, + { + "epoch": 14.591054313099042, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 18268 + }, + { + "epoch": 14.59185303514377, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 18269 + }, + { + "epoch": 14.592651757188499, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 18270 + }, + { + "epoch": 14.593450479233226, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18271 + }, + { + "epoch": 14.594249201277956, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 18272 + }, + { + "epoch": 14.595047923322683, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 18273 + }, + { + "epoch": 14.595846645367413, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 18274 + }, + { + "epoch": 14.59664536741214, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 18275 + }, + { + "epoch": 14.59744408945687, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 18276 + }, + { + "epoch": 14.598242811501597, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 18277 + }, + { + "epoch": 14.599041533546325, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 18278 + }, + { + "epoch": 14.599840255591054, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 18279 + }, + { + "epoch": 14.600638977635782, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 18280 + }, + { + "epoch": 14.601437699680512, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 18281 + }, + { + "epoch": 14.60223642172524, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 18282 + }, + { + "epoch": 14.603035143769969, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 18283 + }, + { + "epoch": 14.603833865814696, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18284 + }, + { + "epoch": 14.604632587859426, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 18285 + }, + { + "epoch": 14.605431309904153, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 18286 + }, + { + "epoch": 14.606230031948883, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 18287 + }, + { + "epoch": 14.60702875399361, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 18288 + }, + { + "epoch": 14.607827476038338, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 18289 + }, + { + "epoch": 14.608626198083067, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 18290 + }, + { + "epoch": 14.609424920127795, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 18291 + }, + { + "epoch": 14.610223642172524, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 18292 + }, + { + "epoch": 14.611022364217252, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 18293 + }, + { + "epoch": 14.611821086261982, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 18294 + }, + { + "epoch": 14.61261980830671, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 18295 + }, + { + "epoch": 14.613418530351439, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 18296 + }, + { + "epoch": 14.614217252396166, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 18297 + }, + { + "epoch": 14.615015974440894, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 18298 + }, + { + "epoch": 14.615814696485623, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 18299 + }, + { + "epoch": 14.616613418530351, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 18300 + }, + { + "epoch": 14.61741214057508, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 18301 + }, + { + "epoch": 14.618210862619808, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 18302 + }, + { + "epoch": 14.619009584664537, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 18303 + }, + { + "epoch": 14.619808306709265, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 18304 + }, + { + "epoch": 14.620607028753994, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 18305 + }, + { + "epoch": 14.621405750798722, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18306 + }, + { + "epoch": 14.62220447284345, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 18307 + }, + { + "epoch": 14.62300319488818, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 18308 + }, + { + "epoch": 14.623801916932907, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18309 + }, + { + "epoch": 14.624600638977636, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 18310 + }, + { + "epoch": 14.625399361022364, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18311 + }, + { + "epoch": 14.626198083067093, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 18312 + }, + { + "epoch": 14.62699680511182, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18313 + }, + { + "epoch": 14.62779552715655, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 18314 + }, + { + "epoch": 14.628594249201278, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 18315 + }, + { + "epoch": 14.629392971246006, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 18316 + }, + { + "epoch": 14.630191693290735, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 18317 + }, + { + "epoch": 14.630990415335463, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 18318 + }, + { + "epoch": 14.631789137380192, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 18319 + }, + { + "epoch": 14.63258785942492, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 18320 + }, + { + "epoch": 14.633386581469649, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18321 + }, + { + "epoch": 14.634185303514377, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 18322 + }, + { + "epoch": 14.634984025559106, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 18323 + }, + { + "epoch": 14.635782747603834, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 18324 + }, + { + "epoch": 14.636581469648561, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 18325 + }, + { + "epoch": 14.63738019169329, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 18326 + }, + { + "epoch": 14.638178913738018, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 18327 + }, + { + "epoch": 14.638977635782748, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 18328 + }, + { + "epoch": 14.639776357827476, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 18329 + }, + { + "epoch": 14.640575079872205, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 18330 + }, + { + "epoch": 14.641373801916933, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 18331 + }, + { + "epoch": 14.642172523961662, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18332 + }, + { + "epoch": 14.64297124600639, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18333 + }, + { + "epoch": 14.643769968051117, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 18334 + }, + { + "epoch": 14.644568690095847, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 18335 + }, + { + "epoch": 14.645367412140574, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 18336 + }, + { + "epoch": 14.646166134185304, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 18337 + }, + { + "epoch": 14.646964856230031, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 18338 + }, + { + "epoch": 14.64776357827476, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 18339 + }, + { + "epoch": 14.648562300319488, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 18340 + }, + { + "epoch": 14.649361022364218, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 18341 + }, + { + "epoch": 14.650159744408946, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 18342 + }, + { + "epoch": 14.650958466453675, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 18343 + }, + { + "epoch": 14.651757188498403, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 18344 + }, + { + "epoch": 14.65255591054313, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 18345 + }, + { + "epoch": 14.65335463258786, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 18346 + }, + { + "epoch": 14.654153354632587, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18347 + }, + { + "epoch": 14.654952076677317, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 18348 + }, + { + "epoch": 14.655750798722044, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 18349 + }, + { + "epoch": 14.656549520766774, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 18350 + }, + { + "epoch": 14.657348242811501, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18351 + }, + { + "epoch": 14.65814696485623, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 18352 + }, + { + "epoch": 14.658945686900958, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 18353 + }, + { + "epoch": 14.659744408945686, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 18354 + }, + { + "epoch": 14.660543130990416, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 18355 + }, + { + "epoch": 14.661341853035143, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 18356 + }, + { + "epoch": 14.662140575079873, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 18357 + }, + { + "epoch": 14.6629392971246, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 18358 + }, + { + "epoch": 14.66373801916933, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 18359 + }, + { + "epoch": 14.664536741214057, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 18360 + }, + { + "epoch": 14.665335463258787, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 18361 + }, + { + "epoch": 14.666134185303514, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 18362 + }, + { + "epoch": 14.666932907348244, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 18363 + }, + { + "epoch": 14.667731629392971, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 18364 + }, + { + "epoch": 14.668530351437699, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 18365 + }, + { + "epoch": 14.669329073482428, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 18366 + }, + { + "epoch": 14.670127795527156, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 18367 + }, + { + "epoch": 14.670926517571885, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 18368 + }, + { + "epoch": 14.671725239616613, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 18369 + }, + { + "epoch": 14.672523961661343, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 18370 + }, + { + "epoch": 14.67332268370607, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 18371 + }, + { + "epoch": 14.6741214057508, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 18372 + }, + { + "epoch": 14.674920127795527, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 18373 + }, + { + "epoch": 14.675718849840255, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 18374 + }, + { + "epoch": 14.676517571884984, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 18375 + }, + { + "epoch": 14.677316293929712, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18376 + }, + { + "epoch": 14.678115015974441, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 18377 + }, + { + "epoch": 14.678913738019169, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 18378 + }, + { + "epoch": 14.679712460063898, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 18379 + }, + { + "epoch": 14.680511182108626, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 18380 + }, + { + "epoch": 14.681309904153355, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 18381 + }, + { + "epoch": 14.682108626198083, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 18382 + }, + { + "epoch": 14.68290734824281, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18383 + }, + { + "epoch": 14.68370607028754, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 18384 + }, + { + "epoch": 14.684504792332268, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 18385 + }, + { + "epoch": 14.685303514376997, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 18386 + }, + { + "epoch": 14.686102236421725, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 18387 + }, + { + "epoch": 14.686900958466454, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 18388 + }, + { + "epoch": 14.687699680511182, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 18389 + }, + { + "epoch": 14.688498402555911, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 18390 + }, + { + "epoch": 14.689297124600639, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 18391 + }, + { + "epoch": 14.690095846645367, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 18392 + }, + { + "epoch": 14.690894568690096, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 18393 + }, + { + "epoch": 14.691693290734824, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 18394 + }, + { + "epoch": 14.692492012779553, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 18395 + }, + { + "epoch": 14.69329073482428, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 18396 + }, + { + "epoch": 14.69408945686901, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 18397 + }, + { + "epoch": 14.694888178913738, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18398 + }, + { + "epoch": 14.695686900958467, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 18399 + }, + { + "epoch": 14.696485623003195, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 18400 + }, + { + "epoch": 14.697284345047922, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18401 + }, + { + "epoch": 14.698083067092652, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 18402 + }, + { + "epoch": 14.69888178913738, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18403 + }, + { + "epoch": 14.699680511182109, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 18404 + }, + { + "epoch": 14.700479233226837, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 18405 + }, + { + "epoch": 14.701277955271566, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18406 + }, + { + "epoch": 14.702076677316294, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 18407 + }, + { + "epoch": 14.702875399361023, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 18408 + }, + { + "epoch": 14.70367412140575, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 18409 + }, + { + "epoch": 14.704472843450478, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 18410 + }, + { + "epoch": 14.705271565495208, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 18411 + }, + { + "epoch": 14.706070287539935, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 18412 + }, + { + "epoch": 14.706869009584665, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 18413 + }, + { + "epoch": 14.707667731629392, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 18414 + }, + { + "epoch": 14.708466453674122, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 18415 + }, + { + "epoch": 14.70926517571885, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 18416 + }, + { + "epoch": 14.710063897763579, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 18417 + }, + { + "epoch": 14.710862619808307, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 18418 + }, + { + "epoch": 14.711661341853034, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 18419 + }, + { + "epoch": 14.712460063897764, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 18420 + }, + { + "epoch": 14.713258785942491, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 18421 + }, + { + "epoch": 14.71405750798722, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 18422 + }, + { + "epoch": 14.714856230031948, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18423 + }, + { + "epoch": 14.715654952076678, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18424 + }, + { + "epoch": 14.716453674121405, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 18425 + }, + { + "epoch": 14.717252396166135, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 18426 + }, + { + "epoch": 14.718051118210862, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 18427 + }, + { + "epoch": 14.718849840255592, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 18428 + }, + { + "epoch": 14.71964856230032, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 18429 + }, + { + "epoch": 14.720447284345047, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 18430 + }, + { + "epoch": 14.721246006389777, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 18431 + }, + { + "epoch": 14.722044728434504, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 18432 + }, + { + "epoch": 14.722843450479234, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 18433 + }, + { + "epoch": 14.723642172523961, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 18434 + }, + { + "epoch": 14.72444089456869, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 18435 + }, + { + "epoch": 14.725239616613418, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 18436 + }, + { + "epoch": 14.726038338658148, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 18437 + }, + { + "epoch": 14.726837060702875, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 18438 + }, + { + "epoch": 14.727635782747605, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 18439 + }, + { + "epoch": 14.728434504792332, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 18440 + }, + { + "epoch": 14.72923322683706, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 18441 + }, + { + "epoch": 14.73003194888179, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 18442 + }, + { + "epoch": 14.730830670926517, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 18443 + }, + { + "epoch": 14.731629392971247, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 18444 + }, + { + "epoch": 14.732428115015974, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 18445 + }, + { + "epoch": 14.733226837060704, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 18446 + }, + { + "epoch": 14.734025559105431, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 18447 + }, + { + "epoch": 14.73482428115016, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 18448 + }, + { + "epoch": 14.735623003194888, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 18449 + }, + { + "epoch": 14.736421725239616, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 18450 + }, + { + "epoch": 14.737220447284345, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 18451 + }, + { + "epoch": 14.738019169329073, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 18452 + }, + { + "epoch": 14.738817891373802, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 18453 + }, + { + "epoch": 14.73961661341853, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 18454 + }, + { + "epoch": 14.74041533546326, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 18455 + }, + { + "epoch": 14.741214057507987, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 18456 + }, + { + "epoch": 14.742012779552716, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 18457 + }, + { + "epoch": 14.742811501597444, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 18458 + }, + { + "epoch": 14.743610223642172, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 18459 + }, + { + "epoch": 14.744408945686901, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 18460 + }, + { + "epoch": 14.745207667731629, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 18461 + }, + { + "epoch": 14.746006389776358, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 18462 + }, + { + "epoch": 14.746805111821086, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 18463 + }, + { + "epoch": 14.747603833865815, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 18464 + }, + { + "epoch": 14.748402555910543, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 18465 + }, + { + "epoch": 14.749201277955272, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 18466 + }, + { + "epoch": 14.75, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18467 + }, + { + "epoch": 14.750798722044728, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 18468 + }, + { + "epoch": 14.751597444089457, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 18469 + }, + { + "epoch": 14.752396166134185, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 18470 + }, + { + "epoch": 14.753194888178914, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 18471 + }, + { + "epoch": 14.753993610223642, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 18472 + }, + { + "epoch": 14.754792332268371, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 18473 + }, + { + "epoch": 14.755591054313099, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 18474 + }, + { + "epoch": 14.756389776357828, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 18475 + }, + { + "epoch": 14.757188498402556, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 18476 + }, + { + "epoch": 14.757987220447284, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 18477 + }, + { + "epoch": 14.758785942492013, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 18478 + }, + { + "epoch": 14.75958466453674, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 18479 + }, + { + "epoch": 14.76038338658147, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 18480 + }, + { + "epoch": 14.761182108626198, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18481 + }, + { + "epoch": 14.761980830670927, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 18482 + }, + { + "epoch": 14.762779552715655, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 18483 + }, + { + "epoch": 14.763578274760384, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 18484 + }, + { + "epoch": 14.764376996805112, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 18485 + }, + { + "epoch": 14.76517571884984, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 18486 + }, + { + "epoch": 14.765974440894569, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 18487 + }, + { + "epoch": 14.766773162939296, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 18488 + }, + { + "epoch": 14.767571884984026, + "grad_norm": 0.337890625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 18489 + }, + { + "epoch": 14.768370607028753, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 18490 + }, + { + "epoch": 14.769169329073483, + "grad_norm": 0.345703125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 18491 + }, + { + "epoch": 14.76996805111821, + "grad_norm": 0.36328125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18492 + }, + { + "epoch": 14.77076677316294, + "grad_norm": 0.40625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 18493 + }, + { + "epoch": 14.771565495207668, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 18494 + }, + { + "epoch": 14.772364217252395, + "grad_norm": 0.3984375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 18495 + }, + { + "epoch": 14.773162939297125, + "grad_norm": 0.34765625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 18496 + }, + { + "epoch": 14.773961661341852, + "grad_norm": 0.42578125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 18497 + }, + { + "epoch": 14.774760383386582, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 18498 + }, + { + "epoch": 14.77555910543131, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 18499 + }, + { + "epoch": 14.776357827476039, + "grad_norm": 0.337890625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 18500 + }, + { + "epoch": 14.777156549520766, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18501 + }, + { + "epoch": 14.777955271565496, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.1117, + "step": 18502 + }, + { + "epoch": 14.778753993610223, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 18503 + }, + { + "epoch": 14.779552715654953, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 18504 + }, + { + "epoch": 14.78035143769968, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 18505 + }, + { + "epoch": 14.781150159744408, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 18506 + }, + { + "epoch": 14.781948881789138, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 18507 + }, + { + "epoch": 14.782747603833865, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 18508 + }, + { + "epoch": 14.783546325878595, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 18509 + }, + { + "epoch": 14.784345047923322, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 18510 + }, + { + "epoch": 14.785143769968052, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 18511 + }, + { + "epoch": 14.78594249201278, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 18512 + }, + { + "epoch": 14.786741214057509, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 18513 + }, + { + "epoch": 14.787539936102236, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 18514 + }, + { + "epoch": 14.788338658146966, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 18515 + }, + { + "epoch": 14.789137380191693, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 18516 + }, + { + "epoch": 14.789936102236421, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 18517 + }, + { + "epoch": 14.79073482428115, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 18518 + }, + { + "epoch": 14.791533546325878, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 18519 + }, + { + "epoch": 14.792332268370608, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 18520 + }, + { + "epoch": 14.793130990415335, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 18521 + }, + { + "epoch": 14.793929712460065, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 18522 + }, + { + "epoch": 14.794728434504792, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 18523 + }, + { + "epoch": 14.795527156549522, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18524 + }, + { + "epoch": 14.79632587859425, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 18525 + }, + { + "epoch": 14.797124600638977, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 18526 + }, + { + "epoch": 14.797923322683706, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 18527 + }, + { + "epoch": 14.798722044728434, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 18528 + }, + { + "epoch": 14.799520766773163, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 18529 + }, + { + "epoch": 14.800319488817891, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 18530 + }, + { + "epoch": 14.80111821086262, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18531 + }, + { + "epoch": 14.801916932907348, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 18532 + }, + { + "epoch": 14.802715654952078, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 18533 + }, + { + "epoch": 14.803514376996805, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 18534 + }, + { + "epoch": 14.804313099041533, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 18535 + }, + { + "epoch": 14.805111821086262, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 18536 + }, + { + "epoch": 14.80591054313099, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 18537 + }, + { + "epoch": 14.80670926517572, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 18538 + }, + { + "epoch": 14.807507987220447, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 18539 + }, + { + "epoch": 14.808306709265176, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 18540 + }, + { + "epoch": 14.809105431309904, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18541 + }, + { + "epoch": 14.809904153354633, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 18542 + }, + { + "epoch": 14.810702875399361, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 18543 + }, + { + "epoch": 14.811501597444089, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 18544 + }, + { + "epoch": 14.812300319488818, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 18545 + }, + { + "epoch": 14.813099041533546, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18546 + }, + { + "epoch": 14.813897763578275, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18547 + }, + { + "epoch": 14.814696485623003, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 18548 + }, + { + "epoch": 14.815495207667732, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 18549 + }, + { + "epoch": 14.81629392971246, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 18550 + }, + { + "epoch": 14.81709265175719, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 18551 + }, + { + "epoch": 14.817891373801917, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 18552 + }, + { + "epoch": 14.818690095846645, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 18553 + }, + { + "epoch": 14.819488817891374, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 18554 + }, + { + "epoch": 14.820287539936102, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 18555 + }, + { + "epoch": 14.821086261980831, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 18556 + }, + { + "epoch": 14.821884984025559, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 18557 + }, + { + "epoch": 14.822683706070288, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18558 + }, + { + "epoch": 14.823482428115016, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 18559 + }, + { + "epoch": 14.824281150159745, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 18560 + }, + { + "epoch": 14.825079872204473, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 18561 + }, + { + "epoch": 14.8258785942492, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 18562 + }, + { + "epoch": 14.82667731629393, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 18563 + }, + { + "epoch": 14.827476038338657, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 18564 + }, + { + "epoch": 14.828274760383387, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 18565 + }, + { + "epoch": 14.829073482428115, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 18566 + }, + { + "epoch": 14.829872204472844, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 18567 + }, + { + "epoch": 14.830670926517572, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 18568 + }, + { + "epoch": 14.831469648562301, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 18569 + }, + { + "epoch": 14.832268370607029, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 18570 + }, + { + "epoch": 14.833067092651756, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 18571 + }, + { + "epoch": 14.833865814696486, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 18572 + }, + { + "epoch": 14.834664536741213, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 18573 + }, + { + "epoch": 14.835463258785943, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 18574 + }, + { + "epoch": 14.83626198083067, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 18575 + }, + { + "epoch": 14.8370607028754, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1093, + "step": 18576 + }, + { + "epoch": 14.837859424920127, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 18577 + }, + { + "epoch": 14.838658146964857, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18578 + }, + { + "epoch": 14.839456869009584, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 18579 + }, + { + "epoch": 14.840255591054314, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 18580 + }, + { + "epoch": 14.841054313099042, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 18581 + }, + { + "epoch": 14.84185303514377, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 18582 + }, + { + "epoch": 14.842651757188499, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 18583 + }, + { + "epoch": 14.843450479233226, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 18584 + }, + { + "epoch": 14.844249201277956, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1066, + "step": 18585 + }, + { + "epoch": 14.845047923322683, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 18586 + }, + { + "epoch": 14.845846645367413, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18587 + }, + { + "epoch": 14.84664536741214, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 18588 + }, + { + "epoch": 14.84744408945687, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 18589 + }, + { + "epoch": 14.848242811501597, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 18590 + }, + { + "epoch": 14.849041533546325, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 18591 + }, + { + "epoch": 14.849840255591054, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 18592 + }, + { + "epoch": 14.850638977635782, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 18593 + }, + { + "epoch": 14.851437699680512, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 18594 + }, + { + "epoch": 14.85223642172524, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18595 + }, + { + "epoch": 14.853035143769969, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 18596 + }, + { + "epoch": 14.853833865814696, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 18597 + }, + { + "epoch": 14.854632587859426, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 18598 + }, + { + "epoch": 14.855431309904153, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 18599 + }, + { + "epoch": 14.856230031948883, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 18600 + }, + { + "epoch": 14.85702875399361, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 18601 + }, + { + "epoch": 14.857827476038338, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 18602 + }, + { + "epoch": 14.858626198083067, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 18603 + }, + { + "epoch": 14.859424920127795, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 18604 + }, + { + "epoch": 14.860223642172524, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 18605 + }, + { + "epoch": 14.861022364217252, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 18606 + }, + { + "epoch": 14.861821086261982, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 18607 + }, + { + "epoch": 14.86261980830671, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 18608 + }, + { + "epoch": 14.863418530351439, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 18609 + }, + { + "epoch": 14.864217252396166, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 18610 + }, + { + "epoch": 14.865015974440894, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 18611 + }, + { + "epoch": 14.865814696485623, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 18612 + }, + { + "epoch": 14.866613418530351, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 18613 + }, + { + "epoch": 14.86741214057508, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 18614 + }, + { + "epoch": 14.868210862619808, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 18615 + }, + { + "epoch": 14.869009584664537, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 18616 + }, + { + "epoch": 14.869808306709265, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 18617 + }, + { + "epoch": 14.870607028753994, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 18618 + }, + { + "epoch": 14.871405750798722, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 18619 + }, + { + "epoch": 14.87220447284345, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 18620 + }, + { + "epoch": 14.87300319488818, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 18621 + }, + { + "epoch": 14.873801916932907, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 18622 + }, + { + "epoch": 14.874600638977636, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 18623 + }, + { + "epoch": 14.875399361022364, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 18624 + }, + { + "epoch": 14.876198083067093, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18625 + }, + { + "epoch": 14.87699680511182, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 18626 + }, + { + "epoch": 14.87779552715655, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18627 + }, + { + "epoch": 14.878594249201278, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 18628 + }, + { + "epoch": 14.879392971246006, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 18629 + }, + { + "epoch": 14.880191693290735, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 18630 + }, + { + "epoch": 14.880990415335463, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 18631 + }, + { + "epoch": 14.881789137380192, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 18632 + }, + { + "epoch": 14.88258785942492, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 18633 + }, + { + "epoch": 14.883386581469649, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 18634 + }, + { + "epoch": 14.884185303514377, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 18635 + }, + { + "epoch": 14.884984025559106, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 18636 + }, + { + "epoch": 14.885782747603834, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 18637 + }, + { + "epoch": 14.886581469648561, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 18638 + }, + { + "epoch": 14.88738019169329, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 18639 + }, + { + "epoch": 14.888178913738018, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18640 + }, + { + "epoch": 14.888977635782748, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 18641 + }, + { + "epoch": 14.889776357827476, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 18642 + }, + { + "epoch": 14.890575079872205, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 18643 + }, + { + "epoch": 14.891373801916933, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 18644 + }, + { + "epoch": 14.892172523961662, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 18645 + }, + { + "epoch": 14.89297124600639, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 18646 + }, + { + "epoch": 14.893769968051117, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 18647 + }, + { + "epoch": 14.894568690095847, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 18648 + }, + { + "epoch": 14.895367412140574, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 18649 + }, + { + "epoch": 14.896166134185304, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 18650 + }, + { + "epoch": 14.896964856230031, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 18651 + }, + { + "epoch": 14.89776357827476, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 18652 + }, + { + "epoch": 14.898562300319488, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 18653 + }, + { + "epoch": 14.899361022364218, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18654 + }, + { + "epoch": 14.900159744408946, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18655 + }, + { + "epoch": 14.900958466453675, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18656 + }, + { + "epoch": 14.901757188498403, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 18657 + }, + { + "epoch": 14.90255591054313, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 18658 + }, + { + "epoch": 14.90335463258786, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 18659 + }, + { + "epoch": 14.904153354632587, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 18660 + }, + { + "epoch": 14.904952076677317, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 18661 + }, + { + "epoch": 14.905750798722044, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 18662 + }, + { + "epoch": 14.906549520766774, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 18663 + }, + { + "epoch": 14.907348242811501, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18664 + }, + { + "epoch": 14.90814696485623, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 18665 + }, + { + "epoch": 14.908945686900958, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 18666 + }, + { + "epoch": 14.909744408945686, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18667 + }, + { + "epoch": 14.910543130990416, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 18668 + }, + { + "epoch": 14.911341853035143, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 18669 + }, + { + "epoch": 14.912140575079873, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 18670 + }, + { + "epoch": 14.9129392971246, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 18671 + }, + { + "epoch": 14.91373801916933, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 18672 + }, + { + "epoch": 14.914536741214057, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 18673 + }, + { + "epoch": 14.915335463258787, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 18674 + }, + { + "epoch": 14.916134185303514, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 18675 + }, + { + "epoch": 14.916932907348244, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 18676 + }, + { + "epoch": 14.917731629392971, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 18677 + }, + { + "epoch": 14.918530351437699, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 18678 + }, + { + "epoch": 14.919329073482428, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 18679 + }, + { + "epoch": 14.920127795527156, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 18680 + }, + { + "epoch": 14.920926517571885, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 18681 + }, + { + "epoch": 14.921725239616613, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18682 + }, + { + "epoch": 14.922523961661343, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18683 + }, + { + "epoch": 14.92332268370607, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 18684 + }, + { + "epoch": 14.9241214057508, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18685 + }, + { + "epoch": 14.924920127795527, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 18686 + }, + { + "epoch": 14.925718849840255, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 18687 + }, + { + "epoch": 14.926517571884984, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 18688 + }, + { + "epoch": 14.927316293929712, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 18689 + }, + { + "epoch": 14.928115015974441, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 18690 + }, + { + "epoch": 14.928913738019169, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 18691 + }, + { + "epoch": 14.929712460063898, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18692 + }, + { + "epoch": 14.930511182108626, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 18693 + }, + { + "epoch": 14.931309904153355, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 18694 + }, + { + "epoch": 14.932108626198083, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 18695 + }, + { + "epoch": 14.93290734824281, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 18696 + }, + { + "epoch": 14.93370607028754, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 18697 + }, + { + "epoch": 14.934504792332268, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 18698 + }, + { + "epoch": 14.935303514376997, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 18699 + }, + { + "epoch": 14.936102236421725, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 18700 + }, + { + "epoch": 14.936900958466454, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 18701 + }, + { + "epoch": 14.937699680511182, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 18702 + }, + { + "epoch": 14.938498402555911, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 18703 + }, + { + "epoch": 14.939297124600639, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 18704 + }, + { + "epoch": 14.940095846645367, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 18705 + }, + { + "epoch": 14.940894568690096, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18706 + }, + { + "epoch": 14.941693290734824, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 18707 + }, + { + "epoch": 14.942492012779553, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 18708 + }, + { + "epoch": 14.94329073482428, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 18709 + }, + { + "epoch": 14.94408945686901, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 18710 + }, + { + "epoch": 14.944888178913738, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 18711 + }, + { + "epoch": 14.945686900958467, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 18712 + }, + { + "epoch": 14.946485623003195, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 18713 + }, + { + "epoch": 14.947284345047922, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18714 + }, + { + "epoch": 14.948083067092652, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 18715 + }, + { + "epoch": 14.94888178913738, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 18716 + }, + { + "epoch": 14.949680511182109, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 18717 + }, + { + "epoch": 14.950479233226837, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 18718 + }, + { + "epoch": 14.951277955271566, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 18719 + }, + { + "epoch": 14.952076677316294, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 18720 + }, + { + "epoch": 14.952875399361023, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 18721 + }, + { + "epoch": 14.95367412140575, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 18722 + }, + { + "epoch": 14.954472843450478, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 18723 + }, + { + "epoch": 14.955271565495208, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 18724 + }, + { + "epoch": 14.956070287539935, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18725 + }, + { + "epoch": 14.956869009584665, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 18726 + }, + { + "epoch": 14.957667731629392, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 18727 + }, + { + "epoch": 14.958466453674122, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 18728 + }, + { + "epoch": 14.95926517571885, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 18729 + }, + { + "epoch": 14.960063897763579, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18730 + }, + { + "epoch": 14.960862619808307, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 18731 + }, + { + "epoch": 14.961661341853034, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 18732 + }, + { + "epoch": 14.962460063897764, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 18733 + }, + { + "epoch": 14.963258785942491, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 18734 + }, + { + "epoch": 14.96405750798722, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 18735 + }, + { + "epoch": 14.964856230031948, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 18736 + }, + { + "epoch": 14.965654952076678, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 18737 + }, + { + "epoch": 14.966453674121405, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18738 + }, + { + "epoch": 14.967252396166135, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.11, + "step": 18739 + }, + { + "epoch": 14.968051118210862, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 18740 + }, + { + "epoch": 14.968849840255592, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 18741 + }, + { + "epoch": 14.96964856230032, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18742 + }, + { + "epoch": 14.970447284345047, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 18743 + }, + { + "epoch": 14.971246006389777, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 18744 + }, + { + "epoch": 14.972044728434504, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 18745 + }, + { + "epoch": 14.972843450479234, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 18746 + }, + { + "epoch": 14.973642172523961, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18747 + }, + { + "epoch": 14.97444089456869, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 18748 + }, + { + "epoch": 14.975239616613418, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 18749 + }, + { + "epoch": 14.976038338658148, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 18750 + }, + { + "epoch": 14.976837060702875, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18751 + }, + { + "epoch": 14.977635782747605, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 18752 + }, + { + "epoch": 14.978434504792332, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 18753 + }, + { + "epoch": 14.97923322683706, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 18754 + }, + { + "epoch": 14.98003194888179, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 18755 + }, + { + "epoch": 14.980830670926517, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 18756 + }, + { + "epoch": 14.981629392971247, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 18757 + }, + { + "epoch": 14.982428115015974, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 18758 + }, + { + "epoch": 14.983226837060704, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 18759 + }, + { + "epoch": 14.984025559105431, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 18760 + }, + { + "epoch": 14.98482428115016, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 18761 + }, + { + "epoch": 14.985623003194888, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 18762 + }, + { + "epoch": 14.986421725239616, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 18763 + }, + { + "epoch": 14.987220447284345, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18764 + }, + { + "epoch": 14.988019169329073, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 18765 + }, + { + "epoch": 14.988817891373802, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 18766 + }, + { + "epoch": 14.98961661341853, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 18767 + }, + { + "epoch": 14.99041533546326, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 18768 + }, + { + "epoch": 14.991214057507987, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 18769 + }, + { + "epoch": 14.992012779552716, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 18770 + }, + { + "epoch": 14.992811501597444, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 18771 + }, + { + "epoch": 14.993610223642172, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 18772 + }, + { + "epoch": 14.994408945686901, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 18773 + }, + { + "epoch": 14.995207667731629, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 18774 + }, + { + "epoch": 14.996006389776358, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 18775 + }, + { + "epoch": 14.996805111821086, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 18776 + }, + { + "epoch": 14.997603833865815, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 18777 + }, + { + "epoch": 14.998402555910543, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 18778 + }, + { + "epoch": 14.999201277955272, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 18779 + }, + { + "epoch": 15.0, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 18780 + }, + { + "epoch": 15.000798722044728, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 18781 + }, + { + "epoch": 15.001597444089457, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 18782 + }, + { + "epoch": 15.002396166134185, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 18783 + }, + { + "epoch": 15.003194888178914, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 18784 + }, + { + "epoch": 15.003993610223642, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 18785 + }, + { + "epoch": 15.004792332268371, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 18786 + }, + { + "epoch": 15.005591054313099, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 18787 + }, + { + "epoch": 15.006389776357828, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 18788 + }, + { + "epoch": 15.007188498402556, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18789 + }, + { + "epoch": 15.007987220447284, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 18790 + }, + { + "epoch": 15.008785942492013, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 18791 + }, + { + "epoch": 15.00958466453674, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18792 + }, + { + "epoch": 15.01038338658147, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 18793 + }, + { + "epoch": 15.011182108626198, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18794 + }, + { + "epoch": 15.011980830670927, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 18795 + }, + { + "epoch": 15.012779552715655, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 18796 + }, + { + "epoch": 15.013578274760384, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 18797 + }, + { + "epoch": 15.014376996805112, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 18798 + }, + { + "epoch": 15.01517571884984, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 18799 + }, + { + "epoch": 15.015974440894569, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 18800 + }, + { + "epoch": 15.016773162939296, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 18801 + }, + { + "epoch": 15.017571884984026, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 18802 + }, + { + "epoch": 15.018370607028753, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 18803 + }, + { + "epoch": 15.019169329073483, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 18804 + }, + { + "epoch": 15.01996805111821, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 18805 + }, + { + "epoch": 15.02076677316294, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 18806 + }, + { + "epoch": 15.021565495207668, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 18807 + }, + { + "epoch": 15.022364217252397, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 18808 + }, + { + "epoch": 15.023162939297125, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 18809 + }, + { + "epoch": 15.023961661341852, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 18810 + }, + { + "epoch": 15.024760383386582, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 18811 + }, + { + "epoch": 15.02555910543131, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 18812 + }, + { + "epoch": 15.026357827476039, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 18813 + }, + { + "epoch": 15.027156549520766, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 18814 + }, + { + "epoch": 15.027955271565496, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 18815 + }, + { + "epoch": 15.028753993610223, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 18816 + }, + { + "epoch": 15.029552715654953, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 18817 + }, + { + "epoch": 15.03035143769968, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 18818 + }, + { + "epoch": 15.031150159744408, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 18819 + }, + { + "epoch": 15.031948881789138, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18820 + }, + { + "epoch": 15.032747603833865, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18821 + }, + { + "epoch": 15.033546325878595, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 18822 + }, + { + "epoch": 15.034345047923322, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 18823 + }, + { + "epoch": 15.035143769968052, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 18824 + }, + { + "epoch": 15.03594249201278, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 18825 + }, + { + "epoch": 15.036741214057509, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 18826 + }, + { + "epoch": 15.037539936102236, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 18827 + }, + { + "epoch": 15.038338658146964, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18828 + }, + { + "epoch": 15.039137380191693, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 18829 + }, + { + "epoch": 15.039936102236421, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 18830 + }, + { + "epoch": 15.04073482428115, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 18831 + }, + { + "epoch": 15.041533546325878, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 18832 + }, + { + "epoch": 15.042332268370608, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 18833 + }, + { + "epoch": 15.043130990415335, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 18834 + }, + { + "epoch": 15.043929712460065, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 18835 + }, + { + "epoch": 15.044728434504792, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18836 + }, + { + "epoch": 15.04552715654952, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 18837 + }, + { + "epoch": 15.04632587859425, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 18838 + }, + { + "epoch": 15.047124600638977, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 18839 + }, + { + "epoch": 15.047923322683706, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 18840 + }, + { + "epoch": 15.048722044728434, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 18841 + }, + { + "epoch": 15.049520766773163, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 18842 + }, + { + "epoch": 15.050319488817891, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 18843 + }, + { + "epoch": 15.05111821086262, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 18844 + }, + { + "epoch": 15.051916932907348, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 18845 + }, + { + "epoch": 15.052715654952078, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 18846 + }, + { + "epoch": 15.053514376996805, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 18847 + }, + { + "epoch": 15.054313099041533, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18848 + }, + { + "epoch": 15.055111821086262, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 18849 + }, + { + "epoch": 15.05591054313099, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 18850 + }, + { + "epoch": 15.05670926517572, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 18851 + }, + { + "epoch": 15.057507987220447, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 18852 + }, + { + "epoch": 15.058306709265176, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 18853 + }, + { + "epoch": 15.059105431309904, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 18854 + }, + { + "epoch": 15.059904153354633, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 18855 + }, + { + "epoch": 15.060702875399361, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 18856 + }, + { + "epoch": 15.061501597444089, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 18857 + }, + { + "epoch": 15.062300319488818, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18858 + }, + { + "epoch": 15.063099041533546, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18859 + }, + { + "epoch": 15.063897763578275, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 18860 + }, + { + "epoch": 15.064696485623003, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 18861 + }, + { + "epoch": 15.065495207667732, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 18862 + }, + { + "epoch": 15.06629392971246, + "grad_norm": 0.35546875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 18863 + }, + { + "epoch": 15.06709265175719, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 18864 + }, + { + "epoch": 15.067891373801917, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 18865 + }, + { + "epoch": 15.068690095846645, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 18866 + }, + { + "epoch": 15.069488817891374, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 18867 + }, + { + "epoch": 15.070287539936102, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 18868 + }, + { + "epoch": 15.071086261980831, + "grad_norm": 0.3203125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 18869 + }, + { + "epoch": 15.071884984025559, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 18870 + }, + { + "epoch": 15.072683706070288, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 18871 + }, + { + "epoch": 15.073482428115016, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 18872 + }, + { + "epoch": 15.074281150159745, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 18873 + }, + { + "epoch": 15.075079872204473, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 18874 + }, + { + "epoch": 15.0758785942492, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 18875 + }, + { + "epoch": 15.07667731629393, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 18876 + }, + { + "epoch": 15.077476038338657, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 18877 + }, + { + "epoch": 15.078274760383387, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 18878 + }, + { + "epoch": 15.079073482428115, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 18879 + }, + { + "epoch": 15.079872204472844, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 18880 + }, + { + "epoch": 15.080670926517572, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 18881 + }, + { + "epoch": 15.081469648562301, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 18882 + }, + { + "epoch": 15.082268370607029, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 18883 + }, + { + "epoch": 15.083067092651758, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 18884 + }, + { + "epoch": 15.083865814696486, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 18885 + }, + { + "epoch": 15.084664536741213, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 18886 + }, + { + "epoch": 15.085463258785943, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 18887 + }, + { + "epoch": 15.08626198083067, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 18888 + }, + { + "epoch": 15.0870607028754, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 18889 + }, + { + "epoch": 15.087859424920127, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18890 + }, + { + "epoch": 15.088658146964857, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 18891 + }, + { + "epoch": 15.089456869009584, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 18892 + }, + { + "epoch": 15.090255591054314, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 18893 + }, + { + "epoch": 15.091054313099042, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 18894 + }, + { + "epoch": 15.09185303514377, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 18895 + }, + { + "epoch": 15.092651757188499, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 18896 + }, + { + "epoch": 15.093450479233226, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 18897 + }, + { + "epoch": 15.094249201277956, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 18898 + }, + { + "epoch": 15.095047923322683, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 18899 + }, + { + "epoch": 15.095846645367413, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 18900 + }, + { + "epoch": 15.09664536741214, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 18901 + }, + { + "epoch": 15.09744408945687, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 18902 + }, + { + "epoch": 15.098242811501597, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 18903 + }, + { + "epoch": 15.099041533546325, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 18904 + }, + { + "epoch": 15.099840255591054, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18905 + }, + { + "epoch": 15.100638977635782, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 18906 + }, + { + "epoch": 15.101437699680512, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 18907 + }, + { + "epoch": 15.10223642172524, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 18908 + }, + { + "epoch": 15.103035143769969, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 18909 + }, + { + "epoch": 15.103833865814696, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 18910 + }, + { + "epoch": 15.104632587859426, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 18911 + }, + { + "epoch": 15.105431309904153, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18912 + }, + { + "epoch": 15.106230031948881, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 18913 + }, + { + "epoch": 15.10702875399361, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 18914 + }, + { + "epoch": 15.107827476038338, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 18915 + }, + { + "epoch": 15.108626198083067, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 18916 + }, + { + "epoch": 15.109424920127795, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 18917 + }, + { + "epoch": 15.110223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 18918 + }, + { + "epoch": 15.111022364217252, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 18919 + }, + { + "epoch": 15.111821086261982, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 18920 + }, + { + "epoch": 15.11261980830671, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 18921 + }, + { + "epoch": 15.113418530351439, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 18922 + }, + { + "epoch": 15.114217252396166, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 18923 + }, + { + "epoch": 15.115015974440894, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 18924 + }, + { + "epoch": 15.115814696485623, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 18925 + }, + { + "epoch": 15.116613418530351, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 18926 + }, + { + "epoch": 15.11741214057508, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 18927 + }, + { + "epoch": 15.118210862619808, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 18928 + }, + { + "epoch": 15.119009584664537, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 18929 + }, + { + "epoch": 15.119808306709265, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 18930 + }, + { + "epoch": 15.120607028753994, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 18931 + }, + { + "epoch": 15.121405750798722, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 18932 + }, + { + "epoch": 15.12220447284345, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 18933 + }, + { + "epoch": 15.12300319488818, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 18934 + }, + { + "epoch": 15.123801916932907, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 18935 + }, + { + "epoch": 15.124600638977636, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 18936 + }, + { + "epoch": 15.125399361022364, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 18937 + }, + { + "epoch": 15.126198083067093, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 18938 + }, + { + "epoch": 15.12699680511182, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 18939 + }, + { + "epoch": 15.12779552715655, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 18940 + }, + { + "epoch": 15.128594249201278, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 18941 + }, + { + "epoch": 15.129392971246006, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 18942 + }, + { + "epoch": 15.130191693290735, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 18943 + }, + { + "epoch": 15.130990415335463, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18944 + }, + { + "epoch": 15.131789137380192, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 18945 + }, + { + "epoch": 15.13258785942492, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 18946 + }, + { + "epoch": 15.133386581469649, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 18947 + }, + { + "epoch": 15.134185303514377, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 18948 + }, + { + "epoch": 15.134984025559106, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 18949 + }, + { + "epoch": 15.135782747603834, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 18950 + }, + { + "epoch": 15.136581469648561, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 18951 + }, + { + "epoch": 15.13738019169329, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 18952 + }, + { + "epoch": 15.138178913738018, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 18953 + }, + { + "epoch": 15.138977635782748, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 18954 + }, + { + "epoch": 15.139776357827476, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 18955 + }, + { + "epoch": 15.140575079872205, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 18956 + }, + { + "epoch": 15.141373801916933, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 18957 + }, + { + "epoch": 15.142172523961662, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 18958 + }, + { + "epoch": 15.14297124600639, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 18959 + }, + { + "epoch": 15.143769968051119, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 18960 + }, + { + "epoch": 15.144568690095847, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 18961 + }, + { + "epoch": 15.145367412140574, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 18962 + }, + { + "epoch": 15.146166134185304, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 18963 + }, + { + "epoch": 15.146964856230031, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 18964 + }, + { + "epoch": 15.14776357827476, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 18965 + }, + { + "epoch": 15.148562300319488, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 18966 + }, + { + "epoch": 15.149361022364218, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 18967 + }, + { + "epoch": 15.150159744408946, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 18968 + }, + { + "epoch": 15.150958466453675, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 18969 + }, + { + "epoch": 15.151757188498403, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 18970 + }, + { + "epoch": 15.15255591054313, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 18971 + }, + { + "epoch": 15.15335463258786, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 18972 + }, + { + "epoch": 15.154153354632587, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 18973 + }, + { + "epoch": 15.154952076677317, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 18974 + }, + { + "epoch": 15.155750798722044, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 18975 + }, + { + "epoch": 15.156549520766774, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 18976 + }, + { + "epoch": 15.157348242811501, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 18977 + }, + { + "epoch": 15.15814696485623, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 18978 + }, + { + "epoch": 15.158945686900958, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 18979 + }, + { + "epoch": 15.159744408945686, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 18980 + }, + { + "epoch": 15.160543130990416, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 18981 + }, + { + "epoch": 15.161341853035143, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 18982 + }, + { + "epoch": 15.162140575079873, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 18983 + }, + { + "epoch": 15.1629392971246, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 18984 + }, + { + "epoch": 15.16373801916933, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 18985 + }, + { + "epoch": 15.164536741214057, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 18986 + }, + { + "epoch": 15.165335463258787, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 18987 + }, + { + "epoch": 15.166134185303514, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 18988 + }, + { + "epoch": 15.166932907348242, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 18989 + }, + { + "epoch": 15.167731629392971, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 18990 + }, + { + "epoch": 15.168530351437699, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 18991 + }, + { + "epoch": 15.169329073482428, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 18992 + }, + { + "epoch": 15.170127795527156, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 18993 + }, + { + "epoch": 15.170926517571885, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 18994 + }, + { + "epoch": 15.171725239616613, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 18995 + }, + { + "epoch": 15.172523961661343, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 18996 + }, + { + "epoch": 15.17332268370607, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 18997 + }, + { + "epoch": 15.1741214057508, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 18998 + }, + { + "epoch": 15.174920127795527, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 18999 + }, + { + "epoch": 15.175718849840255, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 19000 + }, + { + "epoch": 15.176517571884984, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 19001 + }, + { + "epoch": 15.177316293929712, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 19002 + }, + { + "epoch": 15.178115015974441, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 19003 + }, + { + "epoch": 15.178913738019169, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 19004 + }, + { + "epoch": 15.179712460063898, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 19005 + }, + { + "epoch": 15.180511182108626, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 19006 + }, + { + "epoch": 15.181309904153355, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 19007 + }, + { + "epoch": 15.182108626198083, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 19008 + }, + { + "epoch": 15.18290734824281, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 19009 + }, + { + "epoch": 15.18370607028754, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 19010 + }, + { + "epoch": 15.184504792332268, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 19011 + }, + { + "epoch": 15.185303514376997, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 19012 + }, + { + "epoch": 15.186102236421725, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 19013 + }, + { + "epoch": 15.186900958466454, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 19014 + }, + { + "epoch": 15.187699680511182, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 19015 + }, + { + "epoch": 15.188498402555911, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 19016 + }, + { + "epoch": 15.189297124600639, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19017 + }, + { + "epoch": 15.190095846645367, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 19018 + }, + { + "epoch": 15.190894568690096, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 19019 + }, + { + "epoch": 15.191693290734824, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 19020 + }, + { + "epoch": 15.192492012779553, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 19021 + }, + { + "epoch": 15.19329073482428, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 19022 + }, + { + "epoch": 15.19408945686901, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 19023 + }, + { + "epoch": 15.194888178913738, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 19024 + }, + { + "epoch": 15.195686900958467, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 19025 + }, + { + "epoch": 15.196485623003195, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 19026 + }, + { + "epoch": 15.197284345047922, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 19027 + }, + { + "epoch": 15.198083067092652, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 19028 + }, + { + "epoch": 15.19888178913738, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 19029 + }, + { + "epoch": 15.199680511182109, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 19030 + }, + { + "epoch": 15.200479233226837, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 19031 + }, + { + "epoch": 15.201277955271566, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 19032 + }, + { + "epoch": 15.202076677316294, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 19033 + }, + { + "epoch": 15.202875399361023, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 19034 + }, + { + "epoch": 15.20367412140575, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 19035 + }, + { + "epoch": 15.204472843450478, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 19036 + }, + { + "epoch": 15.205271565495208, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 19037 + }, + { + "epoch": 15.206070287539935, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 19038 + }, + { + "epoch": 15.206869009584665, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 19039 + }, + { + "epoch": 15.207667731629392, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 19040 + }, + { + "epoch": 15.208466453674122, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 19041 + }, + { + "epoch": 15.20926517571885, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 19042 + }, + { + "epoch": 15.210063897763579, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 19043 + }, + { + "epoch": 15.210862619808307, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 19044 + }, + { + "epoch": 15.211661341853036, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 19045 + }, + { + "epoch": 15.212460063897764, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 19046 + }, + { + "epoch": 15.213258785942491, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19047 + }, + { + "epoch": 15.21405750798722, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 19048 + }, + { + "epoch": 15.214856230031948, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 19049 + }, + { + "epoch": 15.215654952076678, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 19050 + }, + { + "epoch": 15.216453674121405, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 19051 + }, + { + "epoch": 15.217252396166135, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 19052 + }, + { + "epoch": 15.218051118210862, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 19053 + }, + { + "epoch": 15.218849840255592, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 19054 + }, + { + "epoch": 15.21964856230032, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 19055 + }, + { + "epoch": 15.220447284345047, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 19056 + }, + { + "epoch": 15.221246006389777, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 19057 + }, + { + "epoch": 15.222044728434504, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 19058 + }, + { + "epoch": 15.222843450479234, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 19059 + }, + { + "epoch": 15.223642172523961, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 19060 + }, + { + "epoch": 15.22444089456869, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 19061 + }, + { + "epoch": 15.225239616613418, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 19062 + }, + { + "epoch": 15.226038338658148, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 19063 + }, + { + "epoch": 15.226837060702875, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 19064 + }, + { + "epoch": 15.227635782747603, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 19065 + }, + { + "epoch": 15.228434504792332, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 19066 + }, + { + "epoch": 15.22923322683706, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 19067 + }, + { + "epoch": 15.23003194888179, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 19068 + }, + { + "epoch": 15.230830670926517, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 19069 + }, + { + "epoch": 15.231629392971247, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 19070 + }, + { + "epoch": 15.232428115015974, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 19071 + }, + { + "epoch": 15.233226837060704, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 19072 + }, + { + "epoch": 15.234025559105431, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 19073 + }, + { + "epoch": 15.23482428115016, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 19074 + }, + { + "epoch": 15.235623003194888, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 19075 + }, + { + "epoch": 15.236421725239616, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 19076 + }, + { + "epoch": 15.237220447284345, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 19077 + }, + { + "epoch": 15.238019169329073, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 19078 + }, + { + "epoch": 15.238817891373802, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 19079 + }, + { + "epoch": 15.23961661341853, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 19080 + }, + { + "epoch": 15.24041533546326, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 19081 + }, + { + "epoch": 15.241214057507987, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 19082 + }, + { + "epoch": 15.242012779552716, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 19083 + }, + { + "epoch": 15.242811501597444, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 19084 + }, + { + "epoch": 15.243610223642172, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 19085 + }, + { + "epoch": 15.244408945686901, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 19086 + }, + { + "epoch": 15.245207667731629, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 19087 + }, + { + "epoch": 15.246006389776358, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 19088 + }, + { + "epoch": 15.246805111821086, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 19089 + }, + { + "epoch": 15.247603833865815, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 19090 + }, + { + "epoch": 15.248402555910543, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 19091 + }, + { + "epoch": 15.249201277955272, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 19092 + }, + { + "epoch": 15.25, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 19093 + }, + { + "epoch": 15.250798722044728, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 19094 + }, + { + "epoch": 15.251597444089457, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 19095 + }, + { + "epoch": 15.252396166134185, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 19096 + }, + { + "epoch": 15.253194888178914, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 19097 + }, + { + "epoch": 15.253993610223642, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 19098 + }, + { + "epoch": 15.254792332268371, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 19099 + }, + { + "epoch": 15.255591054313099, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 19100 + }, + { + "epoch": 15.256389776357828, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 19101 + }, + { + "epoch": 15.257188498402556, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 19102 + }, + { + "epoch": 15.257987220447284, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 19103 + }, + { + "epoch": 15.258785942492013, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 19104 + }, + { + "epoch": 15.25958466453674, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 19105 + }, + { + "epoch": 15.26038338658147, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 19106 + }, + { + "epoch": 15.261182108626198, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19107 + }, + { + "epoch": 15.261980830670927, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 19108 + }, + { + "epoch": 15.262779552715655, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 19109 + }, + { + "epoch": 15.263578274760384, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 19110 + }, + { + "epoch": 15.264376996805112, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 19111 + }, + { + "epoch": 15.26517571884984, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 19112 + }, + { + "epoch": 15.265974440894569, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 19113 + }, + { + "epoch": 15.266773162939296, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 19114 + }, + { + "epoch": 15.267571884984026, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 19115 + }, + { + "epoch": 15.268370607028753, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 19116 + }, + { + "epoch": 15.269169329073483, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 19117 + }, + { + "epoch": 15.26996805111821, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 19118 + }, + { + "epoch": 15.27076677316294, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 19119 + }, + { + "epoch": 15.271565495207668, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 19120 + }, + { + "epoch": 15.272364217252397, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19121 + }, + { + "epoch": 15.273162939297125, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 19122 + }, + { + "epoch": 15.273961661341852, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 19123 + }, + { + "epoch": 15.274760383386582, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 19124 + }, + { + "epoch": 15.27555910543131, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 19125 + }, + { + "epoch": 15.276357827476039, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 19126 + }, + { + "epoch": 15.277156549520766, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19127 + }, + { + "epoch": 15.277955271565496, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 19128 + }, + { + "epoch": 15.278753993610223, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 19129 + }, + { + "epoch": 15.279552715654953, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 19130 + }, + { + "epoch": 15.28035143769968, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 19131 + }, + { + "epoch": 15.281150159744408, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 19132 + }, + { + "epoch": 15.281948881789138, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 19133 + }, + { + "epoch": 15.282747603833865, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 19134 + }, + { + "epoch": 15.283546325878595, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 19135 + }, + { + "epoch": 15.284345047923322, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 19136 + }, + { + "epoch": 15.285143769968052, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19137 + }, + { + "epoch": 15.28594249201278, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 19138 + }, + { + "epoch": 15.286741214057509, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 19139 + }, + { + "epoch": 15.287539936102236, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 19140 + }, + { + "epoch": 15.288338658146964, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 19141 + }, + { + "epoch": 15.289137380191693, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 19142 + }, + { + "epoch": 15.289936102236421, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 19143 + }, + { + "epoch": 15.29073482428115, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 19144 + }, + { + "epoch": 15.291533546325878, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 19145 + }, + { + "epoch": 15.292332268370608, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 19146 + }, + { + "epoch": 15.293130990415335, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 19147 + }, + { + "epoch": 15.293929712460065, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 19148 + }, + { + "epoch": 15.294728434504792, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 19149 + }, + { + "epoch": 15.295527156549522, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 19150 + }, + { + "epoch": 15.29632587859425, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 19151 + }, + { + "epoch": 15.297124600638977, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 19152 + }, + { + "epoch": 15.297923322683706, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 19153 + }, + { + "epoch": 15.298722044728434, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 19154 + }, + { + "epoch": 15.299520766773163, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 19155 + }, + { + "epoch": 15.300319488817891, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 19156 + }, + { + "epoch": 15.30111821086262, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 19157 + }, + { + "epoch": 15.301916932907348, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 19158 + }, + { + "epoch": 15.302715654952078, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 19159 + }, + { + "epoch": 15.303514376996805, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 19160 + }, + { + "epoch": 15.304313099041533, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 19161 + }, + { + "epoch": 15.305111821086262, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 19162 + }, + { + "epoch": 15.30591054313099, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 19163 + }, + { + "epoch": 15.30670926517572, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 19164 + }, + { + "epoch": 15.307507987220447, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 19165 + }, + { + "epoch": 15.308306709265176, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 19166 + }, + { + "epoch": 15.309105431309904, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19167 + }, + { + "epoch": 15.309904153354633, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 19168 + }, + { + "epoch": 15.310702875399361, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 19169 + }, + { + "epoch": 15.311501597444089, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 19170 + }, + { + "epoch": 15.312300319488818, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 19171 + }, + { + "epoch": 15.313099041533546, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19172 + }, + { + "epoch": 15.313897763578275, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 19173 + }, + { + "epoch": 15.314696485623003, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 19174 + }, + { + "epoch": 15.315495207667732, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 19175 + }, + { + "epoch": 15.31629392971246, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 19176 + }, + { + "epoch": 15.31709265175719, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 19177 + }, + { + "epoch": 15.317891373801917, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 19178 + }, + { + "epoch": 15.318690095846645, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 19179 + }, + { + "epoch": 15.319488817891374, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 19180 + }, + { + "epoch": 15.320287539936102, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 19181 + }, + { + "epoch": 15.321086261980831, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 19182 + }, + { + "epoch": 15.321884984025559, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 19183 + }, + { + "epoch": 15.322683706070288, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 19184 + }, + { + "epoch": 15.323482428115016, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 19185 + }, + { + "epoch": 15.324281150159745, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 19186 + }, + { + "epoch": 15.325079872204473, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 19187 + }, + { + "epoch": 15.3258785942492, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 19188 + }, + { + "epoch": 15.32667731629393, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 19189 + }, + { + "epoch": 15.327476038338657, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 19190 + }, + { + "epoch": 15.328274760383387, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 19191 + }, + { + "epoch": 15.329073482428115, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 19192 + }, + { + "epoch": 15.329872204472844, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19193 + }, + { + "epoch": 15.330670926517572, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 19194 + }, + { + "epoch": 15.331469648562301, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 19195 + }, + { + "epoch": 15.332268370607029, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 19196 + }, + { + "epoch": 15.333067092651758, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 19197 + }, + { + "epoch": 15.333865814696486, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19198 + }, + { + "epoch": 15.334664536741213, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 19199 + }, + { + "epoch": 15.335463258785943, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 19200 + }, + { + "epoch": 15.33626198083067, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19201 + }, + { + "epoch": 15.3370607028754, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 19202 + }, + { + "epoch": 15.337859424920127, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 19203 + }, + { + "epoch": 15.338658146964857, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 19204 + }, + { + "epoch": 15.339456869009584, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 19205 + }, + { + "epoch": 15.340255591054314, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 19206 + }, + { + "epoch": 15.341054313099042, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 19207 + }, + { + "epoch": 15.34185303514377, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 19208 + }, + { + "epoch": 15.342651757188499, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19209 + }, + { + "epoch": 15.343450479233226, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 19210 + }, + { + "epoch": 15.344249201277956, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 19211 + }, + { + "epoch": 15.345047923322683, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 19212 + }, + { + "epoch": 15.345846645367413, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 19213 + }, + { + "epoch": 15.34664536741214, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 19214 + }, + { + "epoch": 15.34744408945687, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 19215 + }, + { + "epoch": 15.348242811501597, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 19216 + }, + { + "epoch": 15.349041533546325, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 19217 + }, + { + "epoch": 15.349840255591054, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 19218 + }, + { + "epoch": 15.350638977635782, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 19219 + }, + { + "epoch": 15.351437699680512, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 19220 + }, + { + "epoch": 15.35223642172524, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 19221 + }, + { + "epoch": 15.353035143769969, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 19222 + }, + { + "epoch": 15.353833865814696, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 19223 + }, + { + "epoch": 15.354632587859426, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 19224 + }, + { + "epoch": 15.355431309904153, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 19225 + }, + { + "epoch": 15.356230031948883, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 19226 + }, + { + "epoch": 15.35702875399361, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 19227 + }, + { + "epoch": 15.357827476038338, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 19228 + }, + { + "epoch": 15.358626198083067, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 19229 + }, + { + "epoch": 15.359424920127795, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19230 + }, + { + "epoch": 15.360223642172524, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 19231 + }, + { + "epoch": 15.361022364217252, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 19232 + }, + { + "epoch": 15.361821086261982, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 19233 + }, + { + "epoch": 15.36261980830671, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 19234 + }, + { + "epoch": 15.363418530351439, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 19235 + }, + { + "epoch": 15.364217252396166, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 19236 + }, + { + "epoch": 15.365015974440894, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 19237 + }, + { + "epoch": 15.365814696485623, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 19238 + }, + { + "epoch": 15.366613418530351, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 19239 + }, + { + "epoch": 15.36741214057508, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 19240 + }, + { + "epoch": 15.368210862619808, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 19241 + }, + { + "epoch": 15.369009584664537, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 19242 + }, + { + "epoch": 15.369808306709265, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 19243 + }, + { + "epoch": 15.370607028753994, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 19244 + }, + { + "epoch": 15.371405750798722, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 19245 + }, + { + "epoch": 15.37220447284345, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 19246 + }, + { + "epoch": 15.37300319488818, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 19247 + }, + { + "epoch": 15.373801916932907, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 19248 + }, + { + "epoch": 15.374600638977636, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 19249 + }, + { + "epoch": 15.375399361022364, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 19250 + }, + { + "epoch": 15.376198083067093, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 19251 + }, + { + "epoch": 15.37699680511182, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19252 + }, + { + "epoch": 15.37779552715655, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 19253 + }, + { + "epoch": 15.378594249201278, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 19254 + }, + { + "epoch": 15.379392971246006, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 19255 + }, + { + "epoch": 15.380191693290735, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 19256 + }, + { + "epoch": 15.380990415335463, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 19257 + }, + { + "epoch": 15.381789137380192, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 19258 + }, + { + "epoch": 15.38258785942492, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 19259 + }, + { + "epoch": 15.383386581469649, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 19260 + }, + { + "epoch": 15.384185303514377, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 19261 + }, + { + "epoch": 15.384984025559106, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 19262 + }, + { + "epoch": 15.385782747603834, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 19263 + }, + { + "epoch": 15.386581469648561, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 19264 + }, + { + "epoch": 15.38738019169329, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 19265 + }, + { + "epoch": 15.388178913738018, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 19266 + }, + { + "epoch": 15.388977635782748, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 19267 + }, + { + "epoch": 15.389776357827476, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 19268 + }, + { + "epoch": 15.390575079872205, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 19269 + }, + { + "epoch": 15.391373801916933, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 19270 + }, + { + "epoch": 15.392172523961662, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 19271 + }, + { + "epoch": 15.39297124600639, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 19272 + }, + { + "epoch": 15.393769968051117, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 19273 + }, + { + "epoch": 15.394568690095847, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 19274 + }, + { + "epoch": 15.395367412140574, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 19275 + }, + { + "epoch": 15.396166134185304, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 19276 + }, + { + "epoch": 15.396964856230031, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 19277 + }, + { + "epoch": 15.39776357827476, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 19278 + }, + { + "epoch": 15.398562300319488, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 19279 + }, + { + "epoch": 15.399361022364218, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 19280 + }, + { + "epoch": 15.400159744408946, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 19281 + }, + { + "epoch": 15.400958466453675, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 19282 + }, + { + "epoch": 15.401757188498403, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 19283 + }, + { + "epoch": 15.40255591054313, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 19284 + }, + { + "epoch": 15.40335463258786, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 19285 + }, + { + "epoch": 15.404153354632587, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 19286 + }, + { + "epoch": 15.404952076677317, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 19287 + }, + { + "epoch": 15.405750798722044, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 19288 + }, + { + "epoch": 15.406549520766774, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 19289 + }, + { + "epoch": 15.407348242811501, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 19290 + }, + { + "epoch": 15.40814696485623, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 19291 + }, + { + "epoch": 15.408945686900958, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19292 + }, + { + "epoch": 15.409744408945686, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 19293 + }, + { + "epoch": 15.410543130990416, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 19294 + }, + { + "epoch": 15.411341853035143, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 19295 + }, + { + "epoch": 15.412140575079873, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 19296 + }, + { + "epoch": 15.4129392971246, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 19297 + }, + { + "epoch": 15.41373801916933, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 19298 + }, + { + "epoch": 15.414536741214057, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19299 + }, + { + "epoch": 15.415335463258787, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 19300 + }, + { + "epoch": 15.416134185303514, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19301 + }, + { + "epoch": 15.416932907348242, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 19302 + }, + { + "epoch": 15.417731629392971, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 19303 + }, + { + "epoch": 15.418530351437699, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 19304 + }, + { + "epoch": 15.419329073482428, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 19305 + }, + { + "epoch": 15.420127795527156, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 19306 + }, + { + "epoch": 15.420926517571885, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 19307 + }, + { + "epoch": 15.421725239616613, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 19308 + }, + { + "epoch": 15.422523961661343, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 19309 + }, + { + "epoch": 15.42332268370607, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 19310 + }, + { + "epoch": 15.4241214057508, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 19311 + }, + { + "epoch": 15.424920127795527, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 19312 + }, + { + "epoch": 15.425718849840255, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 19313 + }, + { + "epoch": 15.426517571884984, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19314 + }, + { + "epoch": 15.427316293929712, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 19315 + }, + { + "epoch": 15.428115015974441, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 19316 + }, + { + "epoch": 15.428913738019169, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 19317 + }, + { + "epoch": 15.429712460063898, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 19318 + }, + { + "epoch": 15.430511182108626, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 19319 + }, + { + "epoch": 15.431309904153355, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 19320 + }, + { + "epoch": 15.432108626198083, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 19321 + }, + { + "epoch": 15.43290734824281, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19322 + }, + { + "epoch": 15.43370607028754, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 19323 + }, + { + "epoch": 15.434504792332268, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 19324 + }, + { + "epoch": 15.435303514376997, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 19325 + }, + { + "epoch": 15.436102236421725, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 19326 + }, + { + "epoch": 15.436900958466454, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 19327 + }, + { + "epoch": 15.437699680511182, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19328 + }, + { + "epoch": 15.438498402555911, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 19329 + }, + { + "epoch": 15.439297124600639, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 19330 + }, + { + "epoch": 15.440095846645367, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 19331 + }, + { + "epoch": 15.440894568690096, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 19332 + }, + { + "epoch": 15.441693290734824, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 19333 + }, + { + "epoch": 15.442492012779553, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 19334 + }, + { + "epoch": 15.44329073482428, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 19335 + }, + { + "epoch": 15.44408945686901, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 19336 + }, + { + "epoch": 15.444888178913738, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 19337 + }, + { + "epoch": 15.445686900958467, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 19338 + }, + { + "epoch": 15.446485623003195, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 19339 + }, + { + "epoch": 15.447284345047922, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 19340 + }, + { + "epoch": 15.448083067092652, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 19341 + }, + { + "epoch": 15.44888178913738, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 19342 + }, + { + "epoch": 15.449680511182109, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 19343 + }, + { + "epoch": 15.450479233226837, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 19344 + }, + { + "epoch": 15.451277955271566, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 19345 + }, + { + "epoch": 15.452076677316294, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 19346 + }, + { + "epoch": 15.452875399361023, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 19347 + }, + { + "epoch": 15.45367412140575, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 19348 + }, + { + "epoch": 15.454472843450478, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 19349 + }, + { + "epoch": 15.455271565495208, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19350 + }, + { + "epoch": 15.456070287539935, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 19351 + }, + { + "epoch": 15.456869009584665, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 19352 + }, + { + "epoch": 15.457667731629392, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 19353 + }, + { + "epoch": 15.458466453674122, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19354 + }, + { + "epoch": 15.45926517571885, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19355 + }, + { + "epoch": 15.460063897763579, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 19356 + }, + { + "epoch": 15.460862619808307, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 19357 + }, + { + "epoch": 15.461661341853036, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 19358 + }, + { + "epoch": 15.462460063897764, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19359 + }, + { + "epoch": 15.463258785942491, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 19360 + }, + { + "epoch": 15.46405750798722, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 19361 + }, + { + "epoch": 15.464856230031948, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 19362 + }, + { + "epoch": 15.465654952076678, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 19363 + }, + { + "epoch": 15.466453674121405, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 19364 + }, + { + "epoch": 15.467252396166135, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19365 + }, + { + "epoch": 15.468051118210862, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 19366 + }, + { + "epoch": 15.468849840255592, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 19367 + }, + { + "epoch": 15.46964856230032, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 19368 + }, + { + "epoch": 15.470447284345047, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 19369 + }, + { + "epoch": 15.471246006389777, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 19370 + }, + { + "epoch": 15.472044728434504, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 19371 + }, + { + "epoch": 15.472843450479234, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19372 + }, + { + "epoch": 15.473642172523961, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 19373 + }, + { + "epoch": 15.47444089456869, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 19374 + }, + { + "epoch": 15.475239616613418, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 19375 + }, + { + "epoch": 15.476038338658148, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 19376 + }, + { + "epoch": 15.476837060702875, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 19377 + }, + { + "epoch": 15.477635782747603, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 19378 + }, + { + "epoch": 15.478434504792332, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 19379 + }, + { + "epoch": 15.47923322683706, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 19380 + }, + { + "epoch": 15.48003194888179, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19381 + }, + { + "epoch": 15.480830670926517, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 19382 + }, + { + "epoch": 15.481629392971247, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 19383 + }, + { + "epoch": 15.482428115015974, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 19384 + }, + { + "epoch": 15.483226837060704, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 19385 + }, + { + "epoch": 15.484025559105431, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 19386 + }, + { + "epoch": 15.48482428115016, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 19387 + }, + { + "epoch": 15.485623003194888, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 19388 + }, + { + "epoch": 15.486421725239616, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 19389 + }, + { + "epoch": 15.487220447284345, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 19390 + }, + { + "epoch": 15.488019169329073, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 19391 + }, + { + "epoch": 15.488817891373802, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 19392 + }, + { + "epoch": 15.48961661341853, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 19393 + }, + { + "epoch": 15.49041533546326, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 19394 + }, + { + "epoch": 15.491214057507987, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 19395 + }, + { + "epoch": 15.492012779552716, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 19396 + }, + { + "epoch": 15.492811501597444, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19397 + }, + { + "epoch": 15.493610223642172, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 19398 + }, + { + "epoch": 15.494408945686901, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 19399 + }, + { + "epoch": 15.495207667731629, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 19400 + }, + { + "epoch": 15.496006389776358, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 19401 + }, + { + "epoch": 15.496805111821086, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 19402 + }, + { + "epoch": 15.497603833865815, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 19403 + }, + { + "epoch": 15.498402555910543, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 19404 + }, + { + "epoch": 15.499201277955272, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 19405 + }, + { + "epoch": 15.5, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 19406 + }, + { + "epoch": 15.500798722044728, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 19407 + }, + { + "epoch": 15.501597444089457, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 19408 + }, + { + "epoch": 15.502396166134185, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 19409 + }, + { + "epoch": 15.503194888178914, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 19410 + }, + { + "epoch": 15.503993610223642, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 19411 + }, + { + "epoch": 15.504792332268371, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 19412 + }, + { + "epoch": 15.505591054313099, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 19413 + }, + { + "epoch": 15.506389776357828, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 19414 + }, + { + "epoch": 15.507188498402556, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 19415 + }, + { + "epoch": 15.507987220447284, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 19416 + }, + { + "epoch": 15.508785942492013, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 19417 + }, + { + "epoch": 15.50958466453674, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 19418 + }, + { + "epoch": 15.51038338658147, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 19419 + }, + { + "epoch": 15.511182108626198, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 19420 + }, + { + "epoch": 15.511980830670927, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 19421 + }, + { + "epoch": 15.512779552715655, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 19422 + }, + { + "epoch": 15.513578274760384, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 19423 + }, + { + "epoch": 15.514376996805112, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 19424 + }, + { + "epoch": 15.51517571884984, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 19425 + }, + { + "epoch": 15.515974440894569, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 19426 + }, + { + "epoch": 15.516773162939296, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 19427 + }, + { + "epoch": 15.517571884984026, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 19428 + }, + { + "epoch": 15.518370607028753, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 19429 + }, + { + "epoch": 15.519169329073483, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 19430 + }, + { + "epoch": 15.51996805111821, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 19431 + }, + { + "epoch": 15.52076677316294, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 19432 + }, + { + "epoch": 15.521565495207668, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 19433 + }, + { + "epoch": 15.522364217252395, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 19434 + }, + { + "epoch": 15.523162939297125, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 19435 + }, + { + "epoch": 15.523961661341852, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 19436 + }, + { + "epoch": 15.524760383386582, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 19437 + }, + { + "epoch": 15.52555910543131, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 19438 + }, + { + "epoch": 15.526357827476039, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 19439 + }, + { + "epoch": 15.527156549520766, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 19440 + }, + { + "epoch": 15.527955271565496, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 19441 + }, + { + "epoch": 15.528753993610223, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 19442 + }, + { + "epoch": 15.529552715654953, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 19443 + }, + { + "epoch": 15.53035143769968, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 19444 + }, + { + "epoch": 15.531150159744408, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 19445 + }, + { + "epoch": 15.531948881789138, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 19446 + }, + { + "epoch": 15.532747603833865, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 19447 + }, + { + "epoch": 15.533546325878595, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 19448 + }, + { + "epoch": 15.534345047923322, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 19449 + }, + { + "epoch": 15.535143769968052, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 19450 + }, + { + "epoch": 15.53594249201278, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 19451 + }, + { + "epoch": 15.536741214057509, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 19452 + }, + { + "epoch": 15.537539936102236, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 19453 + }, + { + "epoch": 15.538338658146966, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 19454 + }, + { + "epoch": 15.539137380191693, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 19455 + }, + { + "epoch": 15.539936102236421, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 19456 + }, + { + "epoch": 15.54073482428115, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 19457 + }, + { + "epoch": 15.541533546325878, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 19458 + }, + { + "epoch": 15.542332268370608, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 19459 + }, + { + "epoch": 15.543130990415335, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 19460 + }, + { + "epoch": 15.543929712460065, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 19461 + }, + { + "epoch": 15.544728434504792, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 19462 + }, + { + "epoch": 15.545527156549522, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 19463 + }, + { + "epoch": 15.54632587859425, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 19464 + }, + { + "epoch": 15.547124600638977, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 19465 + }, + { + "epoch": 15.547923322683706, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 19466 + }, + { + "epoch": 15.548722044728434, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 19467 + }, + { + "epoch": 15.549520766773163, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 19468 + }, + { + "epoch": 15.550319488817891, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 19469 + }, + { + "epoch": 15.55111821086262, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 19470 + }, + { + "epoch": 15.551916932907348, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 19471 + }, + { + "epoch": 15.552715654952078, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 19472 + }, + { + "epoch": 15.553514376996805, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 19473 + }, + { + "epoch": 15.554313099041533, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 19474 + }, + { + "epoch": 15.555111821086262, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 19475 + }, + { + "epoch": 15.55591054313099, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 19476 + }, + { + "epoch": 15.55670926517572, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 19477 + }, + { + "epoch": 15.557507987220447, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 19478 + }, + { + "epoch": 15.558306709265176, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 19479 + }, + { + "epoch": 15.559105431309904, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 19480 + }, + { + "epoch": 15.559904153354633, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 19481 + }, + { + "epoch": 15.560702875399361, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 19482 + }, + { + "epoch": 15.561501597444089, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 19483 + }, + { + "epoch": 15.562300319488818, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 19484 + }, + { + "epoch": 15.563099041533546, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 19485 + }, + { + "epoch": 15.563897763578275, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 19486 + }, + { + "epoch": 15.564696485623003, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 19487 + }, + { + "epoch": 15.565495207667732, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 19488 + }, + { + "epoch": 15.56629392971246, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 19489 + }, + { + "epoch": 15.56709265175719, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 19490 + }, + { + "epoch": 15.567891373801917, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 19491 + }, + { + "epoch": 15.568690095846645, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 19492 + }, + { + "epoch": 15.569488817891374, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 19493 + }, + { + "epoch": 15.570287539936102, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 19494 + }, + { + "epoch": 15.571086261980831, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 19495 + }, + { + "epoch": 15.571884984025559, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 19496 + }, + { + "epoch": 15.572683706070288, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 19497 + }, + { + "epoch": 15.573482428115016, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 19498 + }, + { + "epoch": 15.574281150159745, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19499 + }, + { + "epoch": 15.575079872204473, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 19500 + }, + { + "epoch": 15.5758785942492, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 19501 + }, + { + "epoch": 15.57667731629393, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 19502 + }, + { + "epoch": 15.577476038338657, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 19503 + }, + { + "epoch": 15.578274760383387, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 19504 + }, + { + "epoch": 15.579073482428115, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 19505 + }, + { + "epoch": 15.579872204472844, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 19506 + }, + { + "epoch": 15.580670926517572, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 19507 + }, + { + "epoch": 15.581469648562301, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 19508 + }, + { + "epoch": 15.582268370607029, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 19509 + }, + { + "epoch": 15.583067092651756, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 19510 + }, + { + "epoch": 15.583865814696486, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 19511 + }, + { + "epoch": 15.584664536741213, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 19512 + }, + { + "epoch": 15.585463258785943, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 19513 + }, + { + "epoch": 15.58626198083067, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 19514 + }, + { + "epoch": 15.5870607028754, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 19515 + }, + { + "epoch": 15.587859424920127, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 19516 + }, + { + "epoch": 15.588658146964857, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 19517 + }, + { + "epoch": 15.589456869009584, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 19518 + }, + { + "epoch": 15.590255591054314, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 19519 + }, + { + "epoch": 15.591054313099042, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 19520 + }, + { + "epoch": 15.59185303514377, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 19521 + }, + { + "epoch": 15.592651757188499, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 19522 + }, + { + "epoch": 15.593450479233226, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 19523 + }, + { + "epoch": 15.594249201277956, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 19524 + }, + { + "epoch": 15.595047923322683, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 19525 + }, + { + "epoch": 15.595846645367413, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 19526 + }, + { + "epoch": 15.59664536741214, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 19527 + }, + { + "epoch": 15.59744408945687, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 19528 + }, + { + "epoch": 15.598242811501597, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 19529 + }, + { + "epoch": 15.599041533546325, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 19530 + }, + { + "epoch": 15.599840255591054, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 19531 + }, + { + "epoch": 15.600638977635782, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 19532 + }, + { + "epoch": 15.601437699680512, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 19533 + }, + { + "epoch": 15.60223642172524, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 19534 + }, + { + "epoch": 15.603035143769969, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 19535 + }, + { + "epoch": 15.603833865814696, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 19536 + }, + { + "epoch": 15.604632587859426, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 19537 + }, + { + "epoch": 15.605431309904153, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 19538 + }, + { + "epoch": 15.606230031948883, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19539 + }, + { + "epoch": 15.60702875399361, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 19540 + }, + { + "epoch": 15.607827476038338, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 19541 + }, + { + "epoch": 15.608626198083067, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 19542 + }, + { + "epoch": 15.609424920127795, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19543 + }, + { + "epoch": 15.610223642172524, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 19544 + }, + { + "epoch": 15.611022364217252, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 19545 + }, + { + "epoch": 15.611821086261982, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 19546 + }, + { + "epoch": 15.61261980830671, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 19547 + }, + { + "epoch": 15.613418530351439, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 19548 + }, + { + "epoch": 15.614217252396166, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 19549 + }, + { + "epoch": 15.615015974440894, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 19550 + }, + { + "epoch": 15.615814696485623, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 19551 + }, + { + "epoch": 15.616613418530351, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 19552 + }, + { + "epoch": 15.61741214057508, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 19553 + }, + { + "epoch": 15.618210862619808, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 19554 + }, + { + "epoch": 15.619009584664537, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 19555 + }, + { + "epoch": 15.619808306709265, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 19556 + }, + { + "epoch": 15.620607028753994, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 19557 + }, + { + "epoch": 15.621405750798722, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 19558 + }, + { + "epoch": 15.62220447284345, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 19559 + }, + { + "epoch": 15.62300319488818, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 19560 + }, + { + "epoch": 15.623801916932907, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 19561 + }, + { + "epoch": 15.624600638977636, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 19562 + }, + { + "epoch": 15.625399361022364, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 19563 + }, + { + "epoch": 15.626198083067093, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 19564 + }, + { + "epoch": 15.62699680511182, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19565 + }, + { + "epoch": 15.62779552715655, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 19566 + }, + { + "epoch": 15.628594249201278, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 19567 + }, + { + "epoch": 15.629392971246006, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 19568 + }, + { + "epoch": 15.630191693290735, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 19569 + }, + { + "epoch": 15.630990415335463, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 19570 + }, + { + "epoch": 15.631789137380192, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 19571 + }, + { + "epoch": 15.63258785942492, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 19572 + }, + { + "epoch": 15.633386581469649, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 19573 + }, + { + "epoch": 15.634185303514377, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 19574 + }, + { + "epoch": 15.634984025559106, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 19575 + }, + { + "epoch": 15.635782747603834, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19576 + }, + { + "epoch": 15.636581469648561, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 19577 + }, + { + "epoch": 15.63738019169329, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 19578 + }, + { + "epoch": 15.638178913738018, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 19579 + }, + { + "epoch": 15.638977635782748, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 19580 + }, + { + "epoch": 15.639776357827476, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 19581 + }, + { + "epoch": 15.640575079872205, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 19582 + }, + { + "epoch": 15.641373801916933, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 19583 + }, + { + "epoch": 15.642172523961662, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 19584 + }, + { + "epoch": 15.64297124600639, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 19585 + }, + { + "epoch": 15.643769968051117, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 19586 + }, + { + "epoch": 15.644568690095847, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 19587 + }, + { + "epoch": 15.645367412140574, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 19588 + }, + { + "epoch": 15.646166134185304, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 19589 + }, + { + "epoch": 15.646964856230031, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 19590 + }, + { + "epoch": 15.64776357827476, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 19591 + }, + { + "epoch": 15.648562300319488, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 19592 + }, + { + "epoch": 15.649361022364218, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 19593 + }, + { + "epoch": 15.650159744408946, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 19594 + }, + { + "epoch": 15.650958466453675, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 19595 + }, + { + "epoch": 15.651757188498403, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 19596 + }, + { + "epoch": 15.65255591054313, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 19597 + }, + { + "epoch": 15.65335463258786, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 19598 + }, + { + "epoch": 15.654153354632587, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 19599 + }, + { + "epoch": 15.654952076677317, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 19600 + }, + { + "epoch": 15.655750798722044, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19601 + }, + { + "epoch": 15.656549520766774, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 19602 + }, + { + "epoch": 15.657348242811501, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 19603 + }, + { + "epoch": 15.65814696485623, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 19604 + }, + { + "epoch": 15.658945686900958, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 19605 + }, + { + "epoch": 15.659744408945686, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 19606 + }, + { + "epoch": 15.660543130990416, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 19607 + }, + { + "epoch": 15.661341853035143, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19608 + }, + { + "epoch": 15.662140575079873, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 19609 + }, + { + "epoch": 15.6629392971246, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 19610 + }, + { + "epoch": 15.66373801916933, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19611 + }, + { + "epoch": 15.664536741214057, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 19612 + }, + { + "epoch": 15.665335463258787, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 19613 + }, + { + "epoch": 15.666134185303514, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 19614 + }, + { + "epoch": 15.666932907348244, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 19615 + }, + { + "epoch": 15.667731629392971, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 19616 + }, + { + "epoch": 15.668530351437699, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 19617 + }, + { + "epoch": 15.669329073482428, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 19618 + }, + { + "epoch": 15.670127795527156, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 19619 + }, + { + "epoch": 15.670926517571885, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 19620 + }, + { + "epoch": 15.671725239616613, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 19621 + }, + { + "epoch": 15.672523961661343, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 19622 + }, + { + "epoch": 15.67332268370607, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 19623 + }, + { + "epoch": 15.6741214057508, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 19624 + }, + { + "epoch": 15.674920127795527, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 19625 + }, + { + "epoch": 15.675718849840255, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 19626 + }, + { + "epoch": 15.676517571884984, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 19627 + }, + { + "epoch": 15.677316293929712, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 19628 + }, + { + "epoch": 15.678115015974441, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 19629 + }, + { + "epoch": 15.678913738019169, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 19630 + }, + { + "epoch": 15.679712460063898, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 19631 + }, + { + "epoch": 15.680511182108626, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19632 + }, + { + "epoch": 15.681309904153355, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 19633 + }, + { + "epoch": 15.682108626198083, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 19634 + }, + { + "epoch": 15.68290734824281, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 19635 + }, + { + "epoch": 15.68370607028754, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 19636 + }, + { + "epoch": 15.684504792332268, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 19637 + }, + { + "epoch": 15.685303514376997, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 19638 + }, + { + "epoch": 15.686102236421725, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 19639 + }, + { + "epoch": 15.686900958466454, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 19640 + }, + { + "epoch": 15.687699680511182, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 19641 + }, + { + "epoch": 15.688498402555911, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 19642 + }, + { + "epoch": 15.689297124600639, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 19643 + }, + { + "epoch": 15.690095846645367, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 19644 + }, + { + "epoch": 15.690894568690096, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19645 + }, + { + "epoch": 15.691693290734824, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 19646 + }, + { + "epoch": 15.692492012779553, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 19647 + }, + { + "epoch": 15.69329073482428, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 19648 + }, + { + "epoch": 15.69408945686901, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 19649 + }, + { + "epoch": 15.694888178913738, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 19650 + }, + { + "epoch": 15.695686900958467, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 19651 + }, + { + "epoch": 15.696485623003195, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 19652 + }, + { + "epoch": 15.697284345047922, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 19653 + }, + { + "epoch": 15.698083067092652, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 19654 + }, + { + "epoch": 15.69888178913738, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 19655 + }, + { + "epoch": 15.699680511182109, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 19656 + }, + { + "epoch": 15.700479233226837, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 19657 + }, + { + "epoch": 15.701277955271566, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 19658 + }, + { + "epoch": 15.702076677316294, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 19659 + }, + { + "epoch": 15.702875399361023, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 19660 + }, + { + "epoch": 15.70367412140575, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 19661 + }, + { + "epoch": 15.704472843450478, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 19662 + }, + { + "epoch": 15.705271565495208, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 19663 + }, + { + "epoch": 15.706070287539935, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 19664 + }, + { + "epoch": 15.706869009584665, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 19665 + }, + { + "epoch": 15.707667731629392, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 19666 + }, + { + "epoch": 15.708466453674122, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 19667 + }, + { + "epoch": 15.70926517571885, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 19668 + }, + { + "epoch": 15.710063897763579, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 19669 + }, + { + "epoch": 15.710862619808307, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 19670 + }, + { + "epoch": 15.711661341853034, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 19671 + }, + { + "epoch": 15.712460063897764, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 19672 + }, + { + "epoch": 15.713258785942491, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 19673 + }, + { + "epoch": 15.71405750798722, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 19674 + }, + { + "epoch": 15.714856230031948, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 19675 + }, + { + "epoch": 15.715654952076678, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 19676 + }, + { + "epoch": 15.716453674121405, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 19677 + }, + { + "epoch": 15.717252396166135, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 19678 + }, + { + "epoch": 15.718051118210862, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 19679 + }, + { + "epoch": 15.718849840255592, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 19680 + }, + { + "epoch": 15.71964856230032, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 19681 + }, + { + "epoch": 15.720447284345047, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 19682 + }, + { + "epoch": 15.721246006389777, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 19683 + }, + { + "epoch": 15.722044728434504, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 19684 + }, + { + "epoch": 15.722843450479234, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 19685 + }, + { + "epoch": 15.723642172523961, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 19686 + }, + { + "epoch": 15.72444089456869, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 19687 + }, + { + "epoch": 15.725239616613418, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19688 + }, + { + "epoch": 15.726038338658148, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 19689 + }, + { + "epoch": 15.726837060702875, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 19690 + }, + { + "epoch": 15.727635782747605, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 19691 + }, + { + "epoch": 15.728434504792332, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 19692 + }, + { + "epoch": 15.72923322683706, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 19693 + }, + { + "epoch": 15.73003194888179, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 19694 + }, + { + "epoch": 15.730830670926517, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 19695 + }, + { + "epoch": 15.731629392971247, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 19696 + }, + { + "epoch": 15.732428115015974, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 19697 + }, + { + "epoch": 15.733226837060704, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 19698 + }, + { + "epoch": 15.734025559105431, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 19699 + }, + { + "epoch": 15.73482428115016, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 19700 + }, + { + "epoch": 15.735623003194888, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 19701 + }, + { + "epoch": 15.736421725239616, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 19702 + }, + { + "epoch": 15.737220447284345, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 19703 + }, + { + "epoch": 15.738019169329073, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 19704 + }, + { + "epoch": 15.738817891373802, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 19705 + }, + { + "epoch": 15.73961661341853, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 19706 + }, + { + "epoch": 15.74041533546326, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 19707 + }, + { + "epoch": 15.741214057507987, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 19708 + }, + { + "epoch": 15.742012779552716, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 19709 + }, + { + "epoch": 15.742811501597444, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 19710 + }, + { + "epoch": 15.743610223642172, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 19711 + }, + { + "epoch": 15.744408945686901, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 19712 + }, + { + "epoch": 15.745207667731629, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 19713 + }, + { + "epoch": 15.746006389776358, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 19714 + }, + { + "epoch": 15.746805111821086, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 19715 + }, + { + "epoch": 15.747603833865815, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 19716 + }, + { + "epoch": 15.748402555910543, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 19717 + }, + { + "epoch": 15.749201277955272, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 19718 + }, + { + "epoch": 15.75, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 19719 + }, + { + "epoch": 15.750798722044728, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 19720 + }, + { + "epoch": 15.751597444089457, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19721 + }, + { + "epoch": 15.752396166134185, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 19722 + }, + { + "epoch": 15.753194888178914, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 19723 + }, + { + "epoch": 15.753993610223642, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 19724 + }, + { + "epoch": 15.754792332268371, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 19725 + }, + { + "epoch": 15.755591054313099, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 19726 + }, + { + "epoch": 15.756389776357828, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 19727 + }, + { + "epoch": 15.757188498402556, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 19728 + }, + { + "epoch": 15.757987220447284, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 19729 + }, + { + "epoch": 15.758785942492013, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 19730 + }, + { + "epoch": 15.75958466453674, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19731 + }, + { + "epoch": 15.76038338658147, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 19732 + }, + { + "epoch": 15.761182108626198, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 19733 + }, + { + "epoch": 15.761980830670927, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 19734 + }, + { + "epoch": 15.762779552715655, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 19735 + }, + { + "epoch": 15.763578274760384, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 19736 + }, + { + "epoch": 15.764376996805112, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 19737 + }, + { + "epoch": 15.76517571884984, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19738 + }, + { + "epoch": 15.765974440894569, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 19739 + }, + { + "epoch": 15.766773162939296, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 19740 + }, + { + "epoch": 15.767571884984026, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 19741 + }, + { + "epoch": 15.768370607028753, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 19742 + }, + { + "epoch": 15.769169329073483, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 19743 + }, + { + "epoch": 15.76996805111821, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 19744 + }, + { + "epoch": 15.77076677316294, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 19745 + }, + { + "epoch": 15.771565495207668, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 19746 + }, + { + "epoch": 15.772364217252395, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 19747 + }, + { + "epoch": 15.773162939297125, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 19748 + }, + { + "epoch": 15.773961661341852, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 19749 + }, + { + "epoch": 15.774760383386582, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 19750 + }, + { + "epoch": 15.77555910543131, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 19751 + }, + { + "epoch": 15.776357827476039, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 19752 + }, + { + "epoch": 15.777156549520766, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 19753 + }, + { + "epoch": 15.777955271565496, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 19754 + }, + { + "epoch": 15.778753993610223, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 19755 + }, + { + "epoch": 15.779552715654953, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 19756 + }, + { + "epoch": 15.78035143769968, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 19757 + }, + { + "epoch": 15.781150159744408, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 19758 + }, + { + "epoch": 15.781948881789138, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 19759 + }, + { + "epoch": 15.782747603833865, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 19760 + }, + { + "epoch": 15.783546325878595, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 19761 + }, + { + "epoch": 15.784345047923322, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 19762 + }, + { + "epoch": 15.785143769968052, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 19763 + }, + { + "epoch": 15.78594249201278, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 19764 + }, + { + "epoch": 15.786741214057509, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 19765 + }, + { + "epoch": 15.787539936102236, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 19766 + }, + { + "epoch": 15.788338658146966, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 19767 + }, + { + "epoch": 15.789137380191693, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 19768 + }, + { + "epoch": 15.789936102236421, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 19769 + }, + { + "epoch": 15.79073482428115, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 19770 + }, + { + "epoch": 15.791533546325878, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 19771 + }, + { + "epoch": 15.792332268370608, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 19772 + }, + { + "epoch": 15.793130990415335, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 19773 + }, + { + "epoch": 15.793929712460065, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 19774 + }, + { + "epoch": 15.794728434504792, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 19775 + }, + { + "epoch": 15.795527156549522, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 19776 + }, + { + "epoch": 15.79632587859425, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 19777 + }, + { + "epoch": 15.797124600638977, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 19778 + }, + { + "epoch": 15.797923322683706, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 19779 + }, + { + "epoch": 15.798722044728434, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 19780 + }, + { + "epoch": 15.799520766773163, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 19781 + }, + { + "epoch": 15.800319488817891, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 19782 + }, + { + "epoch": 15.80111821086262, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 19783 + }, + { + "epoch": 15.801916932907348, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 19784 + }, + { + "epoch": 15.802715654952078, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 19785 + }, + { + "epoch": 15.803514376996805, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 19786 + }, + { + "epoch": 15.804313099041533, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 19787 + }, + { + "epoch": 15.805111821086262, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 19788 + }, + { + "epoch": 15.80591054313099, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 19789 + }, + { + "epoch": 15.80670926517572, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 19790 + }, + { + "epoch": 15.807507987220447, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 19791 + }, + { + "epoch": 15.808306709265176, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 19792 + }, + { + "epoch": 15.809105431309904, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 19793 + }, + { + "epoch": 15.809904153354633, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 19794 + }, + { + "epoch": 15.810702875399361, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 19795 + }, + { + "epoch": 15.811501597444089, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 19796 + }, + { + "epoch": 15.812300319488818, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 19797 + }, + { + "epoch": 15.813099041533546, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 19798 + }, + { + "epoch": 15.813897763578275, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 19799 + }, + { + "epoch": 15.814696485623003, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 19800 + }, + { + "epoch": 15.815495207667732, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 19801 + }, + { + "epoch": 15.81629392971246, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 19802 + }, + { + "epoch": 15.81709265175719, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 19803 + }, + { + "epoch": 15.817891373801917, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 19804 + }, + { + "epoch": 15.818690095846645, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 19805 + }, + { + "epoch": 15.819488817891374, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 19806 + }, + { + "epoch": 15.820287539936102, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 19807 + }, + { + "epoch": 15.821086261980831, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 19808 + }, + { + "epoch": 15.821884984025559, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 19809 + }, + { + "epoch": 15.822683706070288, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 19810 + }, + { + "epoch": 15.823482428115016, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 19811 + }, + { + "epoch": 15.824281150159745, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 19812 + }, + { + "epoch": 15.825079872204473, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 19813 + }, + { + "epoch": 15.8258785942492, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 19814 + }, + { + "epoch": 15.82667731629393, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 19815 + }, + { + "epoch": 15.827476038338657, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 19816 + }, + { + "epoch": 15.828274760383387, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 19817 + }, + { + "epoch": 15.829073482428115, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 19818 + }, + { + "epoch": 15.829872204472844, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 19819 + }, + { + "epoch": 15.830670926517572, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 19820 + }, + { + "epoch": 15.831469648562301, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19821 + }, + { + "epoch": 15.832268370607029, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 19822 + }, + { + "epoch": 15.833067092651756, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 19823 + }, + { + "epoch": 15.833865814696486, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 19824 + }, + { + "epoch": 15.834664536741213, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 19825 + }, + { + "epoch": 15.835463258785943, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 19826 + }, + { + "epoch": 15.83626198083067, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 19827 + }, + { + "epoch": 15.8370607028754, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 19828 + }, + { + "epoch": 15.837859424920127, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 19829 + }, + { + "epoch": 15.838658146964857, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 19830 + }, + { + "epoch": 15.839456869009584, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 19831 + }, + { + "epoch": 15.840255591054314, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 19832 + }, + { + "epoch": 15.841054313099042, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 19833 + }, + { + "epoch": 15.84185303514377, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 19834 + }, + { + "epoch": 15.842651757188499, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 19835 + }, + { + "epoch": 15.843450479233226, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 19836 + }, + { + "epoch": 15.844249201277956, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 19837 + }, + { + "epoch": 15.845047923322683, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19838 + }, + { + "epoch": 15.845846645367413, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19839 + }, + { + "epoch": 15.84664536741214, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 19840 + }, + { + "epoch": 15.84744408945687, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 19841 + }, + { + "epoch": 15.848242811501597, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 19842 + }, + { + "epoch": 15.849041533546325, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 19843 + }, + { + "epoch": 15.849840255591054, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 19844 + }, + { + "epoch": 15.850638977635782, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 19845 + }, + { + "epoch": 15.851437699680512, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 19846 + }, + { + "epoch": 15.85223642172524, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 19847 + }, + { + "epoch": 15.853035143769969, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 19848 + }, + { + "epoch": 15.853833865814696, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 19849 + }, + { + "epoch": 15.854632587859426, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 19850 + }, + { + "epoch": 15.855431309904153, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 19851 + }, + { + "epoch": 15.856230031948883, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 19852 + }, + { + "epoch": 15.85702875399361, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19853 + }, + { + "epoch": 15.857827476038338, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 19854 + }, + { + "epoch": 15.858626198083067, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 19855 + }, + { + "epoch": 15.859424920127795, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 19856 + }, + { + "epoch": 15.860223642172524, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 19857 + }, + { + "epoch": 15.861022364217252, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 19858 + }, + { + "epoch": 15.861821086261982, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 19859 + }, + { + "epoch": 15.86261980830671, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 19860 + }, + { + "epoch": 15.863418530351439, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 19861 + }, + { + "epoch": 15.864217252396166, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19862 + }, + { + "epoch": 15.865015974440894, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 19863 + }, + { + "epoch": 15.865814696485623, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 19864 + }, + { + "epoch": 15.866613418530351, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19865 + }, + { + "epoch": 15.86741214057508, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 19866 + }, + { + "epoch": 15.868210862619808, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 19867 + }, + { + "epoch": 15.869009584664537, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 19868 + }, + { + "epoch": 15.869808306709265, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 19869 + }, + { + "epoch": 15.870607028753994, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 19870 + }, + { + "epoch": 15.871405750798722, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 19871 + }, + { + "epoch": 15.87220447284345, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 19872 + }, + { + "epoch": 15.87300319488818, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 19873 + }, + { + "epoch": 15.873801916932907, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 19874 + }, + { + "epoch": 15.874600638977636, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 19875 + }, + { + "epoch": 15.875399361022364, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19876 + }, + { + "epoch": 15.876198083067093, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 19877 + }, + { + "epoch": 15.87699680511182, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 19878 + }, + { + "epoch": 15.87779552715655, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 19879 + }, + { + "epoch": 15.878594249201278, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 19880 + }, + { + "epoch": 15.879392971246006, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 19881 + }, + { + "epoch": 15.880191693290735, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 19882 + }, + { + "epoch": 15.880990415335463, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 19883 + }, + { + "epoch": 15.881789137380192, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 19884 + }, + { + "epoch": 15.88258785942492, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 19885 + }, + { + "epoch": 15.883386581469649, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 19886 + }, + { + "epoch": 15.884185303514377, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19887 + }, + { + "epoch": 15.884984025559106, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 19888 + }, + { + "epoch": 15.885782747603834, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 19889 + }, + { + "epoch": 15.886581469648561, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 19890 + }, + { + "epoch": 15.88738019169329, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 19891 + }, + { + "epoch": 15.888178913738018, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 19892 + }, + { + "epoch": 15.888977635782748, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 19893 + }, + { + "epoch": 15.889776357827476, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 19894 + }, + { + "epoch": 15.890575079872205, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 19895 + }, + { + "epoch": 15.891373801916933, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 19896 + }, + { + "epoch": 15.892172523961662, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 19897 + }, + { + "epoch": 15.89297124600639, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 19898 + }, + { + "epoch": 15.893769968051117, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 19899 + }, + { + "epoch": 15.894568690095847, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 19900 + }, + { + "epoch": 15.895367412140574, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 19901 + }, + { + "epoch": 15.896166134185304, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 19902 + }, + { + "epoch": 15.896964856230031, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 19903 + }, + { + "epoch": 15.89776357827476, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 19904 + }, + { + "epoch": 15.898562300319488, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 19905 + }, + { + "epoch": 15.899361022364218, + "grad_norm": 0.09619140625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 19906 + }, + { + "epoch": 15.900159744408946, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 19907 + }, + { + "epoch": 15.900958466453675, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 19908 + }, + { + "epoch": 15.901757188498403, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 19909 + }, + { + "epoch": 15.90255591054313, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 19910 + }, + { + "epoch": 15.90335463258786, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 19911 + }, + { + "epoch": 15.904153354632587, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 19912 + }, + { + "epoch": 15.904952076677317, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 19913 + }, + { + "epoch": 15.905750798722044, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 19914 + }, + { + "epoch": 15.906549520766774, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 19915 + }, + { + "epoch": 15.907348242811501, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 19916 + }, + { + "epoch": 15.90814696485623, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 19917 + }, + { + "epoch": 15.908945686900958, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 19918 + }, + { + "epoch": 15.909744408945686, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 19919 + }, + { + "epoch": 15.910543130990416, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 19920 + }, + { + "epoch": 15.911341853035143, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 19921 + }, + { + "epoch": 15.912140575079873, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 19922 + }, + { + "epoch": 15.9129392971246, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 19923 + }, + { + "epoch": 15.91373801916933, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 19924 + }, + { + "epoch": 15.914536741214057, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19925 + }, + { + "epoch": 15.915335463258787, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 19926 + }, + { + "epoch": 15.916134185303514, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 19927 + }, + { + "epoch": 15.916932907348244, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 19928 + }, + { + "epoch": 15.917731629392971, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 19929 + }, + { + "epoch": 15.918530351437699, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 19930 + }, + { + "epoch": 15.919329073482428, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 19931 + }, + { + "epoch": 15.920127795527156, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 19932 + }, + { + "epoch": 15.920926517571885, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 19933 + }, + { + "epoch": 15.921725239616613, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 19934 + }, + { + "epoch": 15.922523961661343, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 19935 + }, + { + "epoch": 15.92332268370607, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 19936 + }, + { + "epoch": 15.9241214057508, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 19937 + }, + { + "epoch": 15.924920127795527, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 19938 + }, + { + "epoch": 15.925718849840255, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 19939 + }, + { + "epoch": 15.926517571884984, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 19940 + }, + { + "epoch": 15.927316293929712, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 19941 + }, + { + "epoch": 15.928115015974441, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 19942 + }, + { + "epoch": 15.928913738019169, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19943 + }, + { + "epoch": 15.929712460063898, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 19944 + }, + { + "epoch": 15.930511182108626, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 19945 + }, + { + "epoch": 15.931309904153355, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 19946 + }, + { + "epoch": 15.932108626198083, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 19947 + }, + { + "epoch": 15.93290734824281, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19948 + }, + { + "epoch": 15.93370607028754, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 19949 + }, + { + "epoch": 15.934504792332268, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 19950 + }, + { + "epoch": 15.935303514376997, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 19951 + }, + { + "epoch": 15.936102236421725, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 19952 + }, + { + "epoch": 15.936900958466454, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 19953 + }, + { + "epoch": 15.937699680511182, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 19954 + }, + { + "epoch": 15.938498402555911, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 19955 + }, + { + "epoch": 15.939297124600639, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 19956 + }, + { + "epoch": 15.940095846645367, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 19957 + }, + { + "epoch": 15.940894568690096, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 19958 + }, + { + "epoch": 15.941693290734824, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 19959 + }, + { + "epoch": 15.942492012779553, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 19960 + }, + { + "epoch": 15.94329073482428, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 19961 + }, + { + "epoch": 15.94408945686901, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 19962 + }, + { + "epoch": 15.944888178913738, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 19963 + }, + { + "epoch": 15.945686900958467, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 19964 + }, + { + "epoch": 15.946485623003195, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 19965 + }, + { + "epoch": 15.947284345047922, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 19966 + }, + { + "epoch": 15.948083067092652, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 19967 + }, + { + "epoch": 15.94888178913738, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 19968 + }, + { + "epoch": 15.949680511182109, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 19969 + }, + { + "epoch": 15.950479233226837, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 19970 + }, + { + "epoch": 15.951277955271566, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 19971 + }, + { + "epoch": 15.952076677316294, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 19972 + }, + { + "epoch": 15.952875399361023, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 19973 + }, + { + "epoch": 15.95367412140575, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19974 + }, + { + "epoch": 15.954472843450478, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 19975 + }, + { + "epoch": 15.955271565495208, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 19976 + }, + { + "epoch": 15.956070287539935, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 19977 + }, + { + "epoch": 15.956869009584665, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 19978 + }, + { + "epoch": 15.957667731629392, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 19979 + }, + { + "epoch": 15.958466453674122, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 19980 + }, + { + "epoch": 15.95926517571885, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 19981 + }, + { + "epoch": 15.960063897763579, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 19982 + }, + { + "epoch": 15.960862619808307, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 19983 + }, + { + "epoch": 15.961661341853034, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 19984 + }, + { + "epoch": 15.962460063897764, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 19985 + }, + { + "epoch": 15.963258785942491, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 19986 + }, + { + "epoch": 15.96405750798722, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 19987 + }, + { + "epoch": 15.964856230031948, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 19988 + }, + { + "epoch": 15.965654952076678, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 19989 + }, + { + "epoch": 15.966453674121405, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 19990 + }, + { + "epoch": 15.967252396166135, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 19991 + }, + { + "epoch": 15.968051118210862, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 19992 + }, + { + "epoch": 15.968849840255592, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 19993 + }, + { + "epoch": 15.96964856230032, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 19994 + }, + { + "epoch": 15.970447284345047, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 19995 + }, + { + "epoch": 15.971246006389777, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 19996 + }, + { + "epoch": 15.972044728434504, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 19997 + }, + { + "epoch": 15.972843450479234, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 19998 + }, + { + "epoch": 15.973642172523961, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 19999 + }, + { + "epoch": 15.97444089456869, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 20000 + }, + { + "epoch": 15.975239616613418, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20001 + }, + { + "epoch": 15.976038338658148, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 20002 + }, + { + "epoch": 15.976837060702875, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 20003 + }, + { + "epoch": 15.977635782747605, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20004 + }, + { + "epoch": 15.978434504792332, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 20005 + }, + { + "epoch": 15.97923322683706, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 20006 + }, + { + "epoch": 15.98003194888179, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 20007 + }, + { + "epoch": 15.980830670926517, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 20008 + }, + { + "epoch": 15.981629392971247, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 20009 + }, + { + "epoch": 15.982428115015974, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 20010 + }, + { + "epoch": 15.983226837060704, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 20011 + }, + { + "epoch": 15.984025559105431, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 20012 + }, + { + "epoch": 15.98482428115016, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20013 + }, + { + "epoch": 15.985623003194888, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 20014 + }, + { + "epoch": 15.986421725239616, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 20015 + }, + { + "epoch": 15.987220447284345, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 20016 + }, + { + "epoch": 15.988019169329073, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 20017 + }, + { + "epoch": 15.988817891373802, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 20018 + }, + { + "epoch": 15.98961661341853, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 20019 + }, + { + "epoch": 15.99041533546326, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 20020 + }, + { + "epoch": 15.991214057507987, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 20021 + }, + { + "epoch": 15.992012779552716, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 20022 + }, + { + "epoch": 15.992811501597444, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 20023 + }, + { + "epoch": 15.993610223642172, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 20024 + }, + { + "epoch": 15.994408945686901, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 20025 + }, + { + "epoch": 15.995207667731629, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 20026 + }, + { + "epoch": 15.996006389776358, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 20027 + }, + { + "epoch": 15.996805111821086, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 20028 + }, + { + "epoch": 15.997603833865815, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 20029 + }, + { + "epoch": 15.998402555910543, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 20030 + }, + { + "epoch": 15.999201277955272, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 20031 + }, + { + "epoch": 16.0, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 20032 + }, + { + "epoch": 16.000798722044728, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 20033 + }, + { + "epoch": 16.001597444089455, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 20034 + }, + { + "epoch": 16.002396166134186, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 20035 + }, + { + "epoch": 16.003194888178914, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 20036 + }, + { + "epoch": 16.00399361022364, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 20037 + }, + { + "epoch": 16.00479233226837, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 20038 + }, + { + "epoch": 16.0055910543131, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 20039 + }, + { + "epoch": 16.00638977635783, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 20040 + }, + { + "epoch": 16.007188498402556, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 20041 + }, + { + "epoch": 16.007987220447284, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20042 + }, + { + "epoch": 16.00878594249201, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 20043 + }, + { + "epoch": 16.009584664536742, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 20044 + }, + { + "epoch": 16.01038338658147, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 20045 + }, + { + "epoch": 16.011182108626198, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 20046 + }, + { + "epoch": 16.011980830670925, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 20047 + }, + { + "epoch": 16.012779552715656, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 20048 + }, + { + "epoch": 16.013578274760384, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 20049 + }, + { + "epoch": 16.01437699680511, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 20050 + }, + { + "epoch": 16.01517571884984, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20051 + }, + { + "epoch": 16.015974440894567, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 20052 + }, + { + "epoch": 16.016773162939298, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 20053 + }, + { + "epoch": 16.017571884984026, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 20054 + }, + { + "epoch": 16.018370607028753, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20055 + }, + { + "epoch": 16.01916932907348, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0749, + "step": 20056 + }, + { + "epoch": 16.019968051118212, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20057 + }, + { + "epoch": 16.02076677316294, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 20058 + }, + { + "epoch": 16.021565495207668, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 20059 + }, + { + "epoch": 16.022364217252395, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 20060 + }, + { + "epoch": 16.023162939297123, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 20061 + }, + { + "epoch": 16.023961661341854, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 20062 + }, + { + "epoch": 16.02476038338658, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 20063 + }, + { + "epoch": 16.02555910543131, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20064 + }, + { + "epoch": 16.026357827476037, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 20065 + }, + { + "epoch": 16.027156549520768, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 20066 + }, + { + "epoch": 16.027955271565496, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20067 + }, + { + "epoch": 16.028753993610223, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1085, + "step": 20068 + }, + { + "epoch": 16.02955271565495, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 20069 + }, + { + "epoch": 16.03035143769968, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 20070 + }, + { + "epoch": 16.03115015974441, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 20071 + }, + { + "epoch": 16.031948881789138, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 20072 + }, + { + "epoch": 16.032747603833865, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 20073 + }, + { + "epoch": 16.033546325878593, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 20074 + }, + { + "epoch": 16.034345047923324, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 20075 + }, + { + "epoch": 16.03514376996805, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 20076 + }, + { + "epoch": 16.03594249201278, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 20077 + }, + { + "epoch": 16.036741214057507, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 20078 + }, + { + "epoch": 16.037539936102238, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20079 + }, + { + "epoch": 16.038338658146966, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 20080 + }, + { + "epoch": 16.039137380191693, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 20081 + }, + { + "epoch": 16.03993610223642, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 20082 + }, + { + "epoch": 16.04073482428115, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 20083 + }, + { + "epoch": 16.04153354632588, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20084 + }, + { + "epoch": 16.042332268370608, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 20085 + }, + { + "epoch": 16.043130990415335, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 20086 + }, + { + "epoch": 16.043929712460063, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 20087 + }, + { + "epoch": 16.044728434504794, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 20088 + }, + { + "epoch": 16.04552715654952, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 20089 + }, + { + "epoch": 16.04632587859425, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 20090 + }, + { + "epoch": 16.047124600638977, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 20091 + }, + { + "epoch": 16.047923322683705, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 20092 + }, + { + "epoch": 16.048722044728436, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 20093 + }, + { + "epoch": 16.049520766773163, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 20094 + }, + { + "epoch": 16.05031948881789, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 20095 + }, + { + "epoch": 16.05111821086262, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20096 + }, + { + "epoch": 16.05191693290735, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 20097 + }, + { + "epoch": 16.052715654952078, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 20098 + }, + { + "epoch": 16.053514376996805, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 20099 + }, + { + "epoch": 16.054313099041533, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 20100 + }, + { + "epoch": 16.05511182108626, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 20101 + }, + { + "epoch": 16.05591054313099, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 20102 + }, + { + "epoch": 16.05670926517572, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 20103 + }, + { + "epoch": 16.057507987220447, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 20104 + }, + { + "epoch": 16.058306709265175, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 20105 + }, + { + "epoch": 16.059105431309906, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 20106 + }, + { + "epoch": 16.059904153354633, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20107 + }, + { + "epoch": 16.06070287539936, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 20108 + }, + { + "epoch": 16.06150159744409, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20109 + }, + { + "epoch": 16.062300319488816, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 20110 + }, + { + "epoch": 16.063099041533548, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 20111 + }, + { + "epoch": 16.063897763578275, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 20112 + }, + { + "epoch": 16.064696485623003, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 20113 + }, + { + "epoch": 16.06549520766773, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 20114 + }, + { + "epoch": 16.06629392971246, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 20115 + }, + { + "epoch": 16.06709265175719, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 20116 + }, + { + "epoch": 16.067891373801917, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20117 + }, + { + "epoch": 16.068690095846645, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 20118 + }, + { + "epoch": 16.069488817891372, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20119 + }, + { + "epoch": 16.070287539936103, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 20120 + }, + { + "epoch": 16.07108626198083, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20121 + }, + { + "epoch": 16.07188498402556, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 20122 + }, + { + "epoch": 16.072683706070286, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 20123 + }, + { + "epoch": 16.073482428115017, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 20124 + }, + { + "epoch": 16.074281150159745, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.105, + "step": 20125 + }, + { + "epoch": 16.075079872204473, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 20126 + }, + { + "epoch": 16.0758785942492, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 20127 + }, + { + "epoch": 16.076677316293928, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20128 + }, + { + "epoch": 16.07747603833866, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 20129 + }, + { + "epoch": 16.078274760383387, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 20130 + }, + { + "epoch": 16.079073482428115, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 20131 + }, + { + "epoch": 16.079872204472842, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 20132 + }, + { + "epoch": 16.080670926517573, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 20133 + }, + { + "epoch": 16.0814696485623, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 20134 + }, + { + "epoch": 16.08226837060703, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 20135 + }, + { + "epoch": 16.083067092651756, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 20136 + }, + { + "epoch": 16.083865814696484, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 20137 + }, + { + "epoch": 16.084664536741215, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 20138 + }, + { + "epoch": 16.085463258785943, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 20139 + }, + { + "epoch": 16.08626198083067, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 20140 + }, + { + "epoch": 16.087060702875398, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 20141 + }, + { + "epoch": 16.08785942492013, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 20142 + }, + { + "epoch": 16.088658146964857, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20143 + }, + { + "epoch": 16.089456869009584, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 20144 + }, + { + "epoch": 16.090255591054312, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 20145 + }, + { + "epoch": 16.09105431309904, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 20146 + }, + { + "epoch": 16.09185303514377, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20147 + }, + { + "epoch": 16.0926517571885, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20148 + }, + { + "epoch": 16.093450479233226, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 20149 + }, + { + "epoch": 16.094249201277954, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 20150 + }, + { + "epoch": 16.095047923322685, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 20151 + }, + { + "epoch": 16.095846645367413, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 20152 + }, + { + "epoch": 16.09664536741214, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 20153 + }, + { + "epoch": 16.097444089456868, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 20154 + }, + { + "epoch": 16.0982428115016, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 20155 + }, + { + "epoch": 16.099041533546327, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 20156 + }, + { + "epoch": 16.099840255591054, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 20157 + }, + { + "epoch": 16.100638977635782, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 20158 + }, + { + "epoch": 16.10143769968051, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 20159 + }, + { + "epoch": 16.10223642172524, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 20160 + }, + { + "epoch": 16.10303514376997, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 20161 + }, + { + "epoch": 16.103833865814696, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20162 + }, + { + "epoch": 16.104632587859424, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 20163 + }, + { + "epoch": 16.105431309904155, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 20164 + }, + { + "epoch": 16.106230031948883, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20165 + }, + { + "epoch": 16.10702875399361, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 20166 + }, + { + "epoch": 16.107827476038338, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20167 + }, + { + "epoch": 16.108626198083066, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 20168 + }, + { + "epoch": 16.109424920127797, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 20169 + }, + { + "epoch": 16.110223642172524, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 20170 + }, + { + "epoch": 16.111022364217252, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20171 + }, + { + "epoch": 16.11182108626198, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 20172 + }, + { + "epoch": 16.11261980830671, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 20173 + }, + { + "epoch": 16.11341853035144, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 20174 + }, + { + "epoch": 16.114217252396166, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20175 + }, + { + "epoch": 16.115015974440894, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 20176 + }, + { + "epoch": 16.11581469648562, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 20177 + }, + { + "epoch": 16.116613418530353, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 20178 + }, + { + "epoch": 16.11741214057508, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 20179 + }, + { + "epoch": 16.118210862619808, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20180 + }, + { + "epoch": 16.119009584664536, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20181 + }, + { + "epoch": 16.119808306709267, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20182 + }, + { + "epoch": 16.120607028753994, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 20183 + }, + { + "epoch": 16.121405750798722, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 20184 + }, + { + "epoch": 16.12220447284345, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 20185 + }, + { + "epoch": 16.123003194888177, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 20186 + }, + { + "epoch": 16.12380191693291, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20187 + }, + { + "epoch": 16.124600638977636, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 20188 + }, + { + "epoch": 16.125399361022364, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 20189 + }, + { + "epoch": 16.12619808306709, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 20190 + }, + { + "epoch": 16.126996805111823, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 20191 + }, + { + "epoch": 16.12779552715655, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 20192 + }, + { + "epoch": 16.128594249201278, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 20193 + }, + { + "epoch": 16.129392971246006, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 20194 + }, + { + "epoch": 16.130191693290733, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 20195 + }, + { + "epoch": 16.130990415335464, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 20196 + }, + { + "epoch": 16.131789137380192, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 20197 + }, + { + "epoch": 16.13258785942492, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 20198 + }, + { + "epoch": 16.133386581469647, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 20199 + }, + { + "epoch": 16.13418530351438, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 20200 + }, + { + "epoch": 16.134984025559106, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20201 + }, + { + "epoch": 16.135782747603834, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 20202 + }, + { + "epoch": 16.13658146964856, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 20203 + }, + { + "epoch": 16.13738019169329, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 20204 + }, + { + "epoch": 16.13817891373802, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 20205 + }, + { + "epoch": 16.138977635782748, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 20206 + }, + { + "epoch": 16.139776357827476, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 20207 + }, + { + "epoch": 16.140575079872203, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1092, + "step": 20208 + }, + { + "epoch": 16.141373801916934, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 20209 + }, + { + "epoch": 16.142172523961662, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 20210 + }, + { + "epoch": 16.14297124600639, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 20211 + }, + { + "epoch": 16.143769968051117, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 20212 + }, + { + "epoch": 16.144568690095845, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 20213 + }, + { + "epoch": 16.145367412140576, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 20214 + }, + { + "epoch": 16.146166134185304, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 20215 + }, + { + "epoch": 16.14696485623003, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 20216 + }, + { + "epoch": 16.14776357827476, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20217 + }, + { + "epoch": 16.14856230031949, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 20218 + }, + { + "epoch": 16.149361022364218, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 20219 + }, + { + "epoch": 16.150159744408946, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 20220 + }, + { + "epoch": 16.150958466453673, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 20221 + }, + { + "epoch": 16.1517571884984, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 20222 + }, + { + "epoch": 16.152555910543132, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 20223 + }, + { + "epoch": 16.15335463258786, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 20224 + }, + { + "epoch": 16.154153354632587, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20225 + }, + { + "epoch": 16.154952076677315, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 20226 + }, + { + "epoch": 16.155750798722046, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 20227 + }, + { + "epoch": 16.156549520766774, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 20228 + }, + { + "epoch": 16.1573482428115, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 20229 + }, + { + "epoch": 16.15814696485623, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 20230 + }, + { + "epoch": 16.15894568690096, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 20231 + }, + { + "epoch": 16.159744408945688, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20232 + }, + { + "epoch": 16.160543130990416, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 20233 + }, + { + "epoch": 16.161341853035143, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 20234 + }, + { + "epoch": 16.16214057507987, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 20235 + }, + { + "epoch": 16.162939297124602, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 20236 + }, + { + "epoch": 16.16373801916933, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 20237 + }, + { + "epoch": 16.164536741214057, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 20238 + }, + { + "epoch": 16.165335463258785, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 20239 + }, + { + "epoch": 16.166134185303516, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 20240 + }, + { + "epoch": 16.166932907348244, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20241 + }, + { + "epoch": 16.16773162939297, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 20242 + }, + { + "epoch": 16.1685303514377, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 20243 + }, + { + "epoch": 16.169329073482427, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 20244 + }, + { + "epoch": 16.170127795527158, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 20245 + }, + { + "epoch": 16.170926517571885, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 20246 + }, + { + "epoch": 16.171725239616613, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20247 + }, + { + "epoch": 16.17252396166134, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 20248 + }, + { + "epoch": 16.173322683706072, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 20249 + }, + { + "epoch": 16.1741214057508, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 20250 + }, + { + "epoch": 16.174920127795527, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 20251 + }, + { + "epoch": 16.175718849840255, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 20252 + }, + { + "epoch": 16.176517571884983, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 20253 + }, + { + "epoch": 16.177316293929714, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 20254 + }, + { + "epoch": 16.17811501597444, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 20255 + }, + { + "epoch": 16.17891373801917, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 20256 + }, + { + "epoch": 16.179712460063897, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 20257 + }, + { + "epoch": 16.180511182108628, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 20258 + }, + { + "epoch": 16.181309904153355, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 20259 + }, + { + "epoch": 16.182108626198083, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 20260 + }, + { + "epoch": 16.18290734824281, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 20261 + }, + { + "epoch": 16.18370607028754, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 20262 + }, + { + "epoch": 16.18450479233227, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20263 + }, + { + "epoch": 16.185303514376997, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 20264 + }, + { + "epoch": 16.186102236421725, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 20265 + }, + { + "epoch": 16.186900958466452, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 20266 + }, + { + "epoch": 16.187699680511184, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 20267 + }, + { + "epoch": 16.18849840255591, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 20268 + }, + { + "epoch": 16.18929712460064, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 20269 + }, + { + "epoch": 16.190095846645367, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 20270 + }, + { + "epoch": 16.190894568690094, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 20271 + }, + { + "epoch": 16.191693290734825, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 20272 + }, + { + "epoch": 16.192492012779553, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 20273 + }, + { + "epoch": 16.19329073482428, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 20274 + }, + { + "epoch": 16.19408945686901, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 20275 + }, + { + "epoch": 16.19488817891374, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 20276 + }, + { + "epoch": 16.195686900958467, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 20277 + }, + { + "epoch": 16.196485623003195, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20278 + }, + { + "epoch": 16.197284345047922, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 20279 + }, + { + "epoch": 16.19808306709265, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 20280 + }, + { + "epoch": 16.19888178913738, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 20281 + }, + { + "epoch": 16.19968051118211, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 20282 + }, + { + "epoch": 16.200479233226837, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 20283 + }, + { + "epoch": 16.201277955271564, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20284 + }, + { + "epoch": 16.202076677316295, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 20285 + }, + { + "epoch": 16.202875399361023, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 20286 + }, + { + "epoch": 16.20367412140575, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 20287 + }, + { + "epoch": 16.20447284345048, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 20288 + }, + { + "epoch": 16.205271565495206, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 20289 + }, + { + "epoch": 16.206070287539937, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 20290 + }, + { + "epoch": 16.206869009584665, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 20291 + }, + { + "epoch": 16.207667731629392, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 20292 + }, + { + "epoch": 16.20846645367412, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 20293 + }, + { + "epoch": 16.20926517571885, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 20294 + }, + { + "epoch": 16.21006389776358, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 20295 + }, + { + "epoch": 16.210862619808307, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 20296 + }, + { + "epoch": 16.211661341853034, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 20297 + }, + { + "epoch": 16.212460063897762, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20298 + }, + { + "epoch": 16.213258785942493, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 20299 + }, + { + "epoch": 16.21405750798722, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 20300 + }, + { + "epoch": 16.21485623003195, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20301 + }, + { + "epoch": 16.215654952076676, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 20302 + }, + { + "epoch": 16.216453674121407, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 20303 + }, + { + "epoch": 16.217252396166135, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 20304 + }, + { + "epoch": 16.218051118210862, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 20305 + }, + { + "epoch": 16.21884984025559, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 20306 + }, + { + "epoch": 16.21964856230032, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 20307 + }, + { + "epoch": 16.22044728434505, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 20308 + }, + { + "epoch": 16.221246006389777, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20309 + }, + { + "epoch": 16.222044728434504, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20310 + }, + { + "epoch": 16.222843450479232, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 20311 + }, + { + "epoch": 16.223642172523963, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 20312 + }, + { + "epoch": 16.22444089456869, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 20313 + }, + { + "epoch": 16.22523961661342, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 20314 + }, + { + "epoch": 16.226038338658146, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 20315 + }, + { + "epoch": 16.226837060702877, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 20316 + }, + { + "epoch": 16.227635782747605, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 20317 + }, + { + "epoch": 16.228434504792332, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 20318 + }, + { + "epoch": 16.22923322683706, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 20319 + }, + { + "epoch": 16.230031948881788, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 20320 + }, + { + "epoch": 16.23083067092652, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 20321 + }, + { + "epoch": 16.231629392971247, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 20322 + }, + { + "epoch": 16.232428115015974, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 20323 + }, + { + "epoch": 16.233226837060702, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 20324 + }, + { + "epoch": 16.234025559105433, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 20325 + }, + { + "epoch": 16.23482428115016, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 20326 + }, + { + "epoch": 16.23562300319489, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 20327 + }, + { + "epoch": 16.236421725239616, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 20328 + }, + { + "epoch": 16.237220447284344, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 20329 + }, + { + "epoch": 16.238019169329075, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 20330 + }, + { + "epoch": 16.238817891373802, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 20331 + }, + { + "epoch": 16.23961661341853, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 20332 + }, + { + "epoch": 16.240415335463258, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 20333 + }, + { + "epoch": 16.24121405750799, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 20334 + }, + { + "epoch": 16.242012779552716, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 20335 + }, + { + "epoch": 16.242811501597444, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 20336 + }, + { + "epoch": 16.24361022364217, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20337 + }, + { + "epoch": 16.2444089456869, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 20338 + }, + { + "epoch": 16.24520766773163, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 20339 + }, + { + "epoch": 16.24600638977636, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 20340 + }, + { + "epoch": 16.246805111821086, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 20341 + }, + { + "epoch": 16.247603833865814, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20342 + }, + { + "epoch": 16.248402555910545, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20343 + }, + { + "epoch": 16.249201277955272, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20344 + }, + { + "epoch": 16.25, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 20345 + }, + { + "epoch": 16.250798722044728, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 20346 + }, + { + "epoch": 16.251597444089455, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 20347 + }, + { + "epoch": 16.252396166134186, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 20348 + }, + { + "epoch": 16.253194888178914, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 20349 + }, + { + "epoch": 16.25399361022364, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 20350 + }, + { + "epoch": 16.25479233226837, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 20351 + }, + { + "epoch": 16.2555910543131, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 20352 + }, + { + "epoch": 16.25638977635783, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 20353 + }, + { + "epoch": 16.257188498402556, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20354 + }, + { + "epoch": 16.257987220447284, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 20355 + }, + { + "epoch": 16.25878594249201, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 20356 + }, + { + "epoch": 16.259584664536742, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 20357 + }, + { + "epoch": 16.26038338658147, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 20358 + }, + { + "epoch": 16.261182108626198, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 20359 + }, + { + "epoch": 16.261980830670925, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 20360 + }, + { + "epoch": 16.262779552715656, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 20361 + }, + { + "epoch": 16.263578274760384, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 20362 + }, + { + "epoch": 16.26437699680511, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 20363 + }, + { + "epoch": 16.26517571884984, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 20364 + }, + { + "epoch": 16.265974440894567, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 20365 + }, + { + "epoch": 16.266773162939298, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 20366 + }, + { + "epoch": 16.267571884984026, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 20367 + }, + { + "epoch": 16.268370607028753, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 20368 + }, + { + "epoch": 16.26916932907348, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20369 + }, + { + "epoch": 16.269968051118212, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 20370 + }, + { + "epoch": 16.27076677316294, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 20371 + }, + { + "epoch": 16.271565495207668, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 20372 + }, + { + "epoch": 16.272364217252395, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 20373 + }, + { + "epoch": 16.273162939297123, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20374 + }, + { + "epoch": 16.273961661341854, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 20375 + }, + { + "epoch": 16.27476038338658, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 20376 + }, + { + "epoch": 16.27555910543131, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 20377 + }, + { + "epoch": 16.276357827476037, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 20378 + }, + { + "epoch": 16.277156549520768, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 20379 + }, + { + "epoch": 16.277955271565496, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 20380 + }, + { + "epoch": 16.278753993610223, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 20381 + }, + { + "epoch": 16.27955271565495, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 20382 + }, + { + "epoch": 16.28035143769968, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 20383 + }, + { + "epoch": 16.28115015974441, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 20384 + }, + { + "epoch": 16.281948881789138, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 20385 + }, + { + "epoch": 16.282747603833865, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 20386 + }, + { + "epoch": 16.283546325878593, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 20387 + }, + { + "epoch": 16.284345047923324, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 20388 + }, + { + "epoch": 16.28514376996805, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20389 + }, + { + "epoch": 16.28594249201278, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 20390 + }, + { + "epoch": 16.286741214057507, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 20391 + }, + { + "epoch": 16.287539936102238, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 20392 + }, + { + "epoch": 16.288338658146966, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 20393 + }, + { + "epoch": 16.289137380191693, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 20394 + }, + { + "epoch": 16.28993610223642, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20395 + }, + { + "epoch": 16.29073482428115, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 20396 + }, + { + "epoch": 16.29153354632588, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 20397 + }, + { + "epoch": 16.292332268370608, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 20398 + }, + { + "epoch": 16.293130990415335, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 20399 + }, + { + "epoch": 16.293929712460063, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 20400 + }, + { + "epoch": 16.294728434504794, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 20401 + }, + { + "epoch": 16.29552715654952, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20402 + }, + { + "epoch": 16.29632587859425, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 20403 + }, + { + "epoch": 16.297124600638977, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20404 + }, + { + "epoch": 16.297923322683705, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 20405 + }, + { + "epoch": 16.298722044728436, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 20406 + }, + { + "epoch": 16.299520766773163, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 20407 + }, + { + "epoch": 16.30031948881789, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 20408 + }, + { + "epoch": 16.30111821086262, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 20409 + }, + { + "epoch": 16.30191693290735, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 20410 + }, + { + "epoch": 16.302715654952078, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 20411 + }, + { + "epoch": 16.303514376996805, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 20412 + }, + { + "epoch": 16.304313099041533, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20413 + }, + { + "epoch": 16.30511182108626, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 20414 + }, + { + "epoch": 16.30591054313099, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 20415 + }, + { + "epoch": 16.30670926517572, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 20416 + }, + { + "epoch": 16.307507987220447, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20417 + }, + { + "epoch": 16.308306709265175, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 20418 + }, + { + "epoch": 16.309105431309906, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 20419 + }, + { + "epoch": 16.309904153354633, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 20420 + }, + { + "epoch": 16.31070287539936, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20421 + }, + { + "epoch": 16.31150159744409, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 20422 + }, + { + "epoch": 16.312300319488816, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 20423 + }, + { + "epoch": 16.313099041533548, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 20424 + }, + { + "epoch": 16.313897763578275, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20425 + }, + { + "epoch": 16.314696485623003, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 20426 + }, + { + "epoch": 16.31549520766773, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 20427 + }, + { + "epoch": 16.31629392971246, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 20428 + }, + { + "epoch": 16.31709265175719, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 20429 + }, + { + "epoch": 16.317891373801917, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 20430 + }, + { + "epoch": 16.318690095846645, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 20431 + }, + { + "epoch": 16.319488817891372, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 20432 + }, + { + "epoch": 16.320287539936103, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 20433 + }, + { + "epoch": 16.32108626198083, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 20434 + }, + { + "epoch": 16.32188498402556, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20435 + }, + { + "epoch": 16.322683706070286, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 20436 + }, + { + "epoch": 16.323482428115017, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 20437 + }, + { + "epoch": 16.324281150159745, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 20438 + }, + { + "epoch": 16.325079872204473, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 20439 + }, + { + "epoch": 16.3258785942492, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 20440 + }, + { + "epoch": 16.326677316293928, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20441 + }, + { + "epoch": 16.32747603833866, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 20442 + }, + { + "epoch": 16.328274760383387, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 20443 + }, + { + "epoch": 16.329073482428115, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 20444 + }, + { + "epoch": 16.329872204472842, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 20445 + }, + { + "epoch": 16.330670926517573, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 20446 + }, + { + "epoch": 16.3314696485623, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 20447 + }, + { + "epoch": 16.33226837060703, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 20448 + }, + { + "epoch": 16.333067092651756, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20449 + }, + { + "epoch": 16.333865814696484, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 20450 + }, + { + "epoch": 16.334664536741215, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 20451 + }, + { + "epoch": 16.335463258785943, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 20452 + }, + { + "epoch": 16.33626198083067, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 20453 + }, + { + "epoch": 16.337060702875398, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 20454 + }, + { + "epoch": 16.33785942492013, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 20455 + }, + { + "epoch": 16.338658146964857, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 20456 + }, + { + "epoch": 16.339456869009584, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 20457 + }, + { + "epoch": 16.340255591054312, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 20458 + }, + { + "epoch": 16.341054313099043, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 20459 + }, + { + "epoch": 16.34185303514377, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 20460 + }, + { + "epoch": 16.3426517571885, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 20461 + }, + { + "epoch": 16.343450479233226, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 20462 + }, + { + "epoch": 16.344249201277954, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 20463 + }, + { + "epoch": 16.345047923322685, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20464 + }, + { + "epoch": 16.345846645367413, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 20465 + }, + { + "epoch": 16.34664536741214, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 20466 + }, + { + "epoch": 16.347444089456868, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 20467 + }, + { + "epoch": 16.3482428115016, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20468 + }, + { + "epoch": 16.349041533546327, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 20469 + }, + { + "epoch": 16.349840255591054, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20470 + }, + { + "epoch": 16.350638977635782, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 20471 + }, + { + "epoch": 16.35143769968051, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20472 + }, + { + "epoch": 16.35223642172524, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20473 + }, + { + "epoch": 16.35303514376997, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20474 + }, + { + "epoch": 16.353833865814696, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 20475 + }, + { + "epoch": 16.354632587859424, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 20476 + }, + { + "epoch": 16.355431309904155, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 20477 + }, + { + "epoch": 16.356230031948883, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 20478 + }, + { + "epoch": 16.35702875399361, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 20479 + }, + { + "epoch": 16.357827476038338, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 20480 + }, + { + "epoch": 16.358626198083066, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 20481 + }, + { + "epoch": 16.359424920127797, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 20482 + }, + { + "epoch": 16.360223642172524, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 20483 + }, + { + "epoch": 16.361022364217252, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 20484 + }, + { + "epoch": 16.36182108626198, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 20485 + }, + { + "epoch": 16.36261980830671, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 20486 + }, + { + "epoch": 16.36341853035144, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 20487 + }, + { + "epoch": 16.364217252396166, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 20488 + }, + { + "epoch": 16.365015974440894, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 20489 + }, + { + "epoch": 16.36581469648562, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 20490 + }, + { + "epoch": 16.366613418530353, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 20491 + }, + { + "epoch": 16.36741214057508, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 20492 + }, + { + "epoch": 16.368210862619808, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 20493 + }, + { + "epoch": 16.369009584664536, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 20494 + }, + { + "epoch": 16.369808306709267, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 20495 + }, + { + "epoch": 16.370607028753994, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20496 + }, + { + "epoch": 16.371405750798722, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 20497 + }, + { + "epoch": 16.37220447284345, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20498 + }, + { + "epoch": 16.373003194888177, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 20499 + }, + { + "epoch": 16.37380191693291, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 20500 + }, + { + "epoch": 16.374600638977636, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 20501 + }, + { + "epoch": 16.375399361022364, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 20502 + }, + { + "epoch": 16.37619808306709, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 20503 + }, + { + "epoch": 16.376996805111823, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 20504 + }, + { + "epoch": 16.37779552715655, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 20505 + }, + { + "epoch": 16.378594249201278, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 20506 + }, + { + "epoch": 16.379392971246006, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 20507 + }, + { + "epoch": 16.380191693290733, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 20508 + }, + { + "epoch": 16.380990415335464, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 20509 + }, + { + "epoch": 16.381789137380192, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20510 + }, + { + "epoch": 16.38258785942492, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20511 + }, + { + "epoch": 16.383386581469647, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20512 + }, + { + "epoch": 16.38418530351438, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 20513 + }, + { + "epoch": 16.384984025559106, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 20514 + }, + { + "epoch": 16.385782747603834, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 20515 + }, + { + "epoch": 16.38658146964856, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 20516 + }, + { + "epoch": 16.38738019169329, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 20517 + }, + { + "epoch": 16.38817891373802, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 20518 + }, + { + "epoch": 16.388977635782748, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 20519 + }, + { + "epoch": 16.389776357827476, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 20520 + }, + { + "epoch": 16.390575079872203, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 20521 + }, + { + "epoch": 16.391373801916934, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 20522 + }, + { + "epoch": 16.392172523961662, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 20523 + }, + { + "epoch": 16.39297124600639, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20524 + }, + { + "epoch": 16.393769968051117, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 20525 + }, + { + "epoch": 16.394568690095845, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 20526 + }, + { + "epoch": 16.395367412140576, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20527 + }, + { + "epoch": 16.396166134185304, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 20528 + }, + { + "epoch": 16.39696485623003, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 20529 + }, + { + "epoch": 16.39776357827476, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 20530 + }, + { + "epoch": 16.39856230031949, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 20531 + }, + { + "epoch": 16.399361022364218, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 20532 + }, + { + "epoch": 16.400159744408946, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 20533 + }, + { + "epoch": 16.400958466453673, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 20534 + }, + { + "epoch": 16.4017571884984, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 20535 + }, + { + "epoch": 16.402555910543132, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 20536 + }, + { + "epoch": 16.40335463258786, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 20537 + }, + { + "epoch": 16.404153354632587, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 20538 + }, + { + "epoch": 16.404952076677315, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20539 + }, + { + "epoch": 16.405750798722046, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 20540 + }, + { + "epoch": 16.406549520766774, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 20541 + }, + { + "epoch": 16.4073482428115, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 20542 + }, + { + "epoch": 16.40814696485623, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 20543 + }, + { + "epoch": 16.408945686900957, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20544 + }, + { + "epoch": 16.409744408945688, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 20545 + }, + { + "epoch": 16.410543130990416, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 20546 + }, + { + "epoch": 16.411341853035143, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 20547 + }, + { + "epoch": 16.41214057507987, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 20548 + }, + { + "epoch": 16.412939297124602, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 20549 + }, + { + "epoch": 16.41373801916933, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 20550 + }, + { + "epoch": 16.414536741214057, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 20551 + }, + { + "epoch": 16.415335463258785, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 20552 + }, + { + "epoch": 16.416134185303516, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 20553 + }, + { + "epoch": 16.416932907348244, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 20554 + }, + { + "epoch": 16.41773162939297, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 20555 + }, + { + "epoch": 16.4185303514377, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 20556 + }, + { + "epoch": 16.419329073482427, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 20557 + }, + { + "epoch": 16.420127795527158, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 20558 + }, + { + "epoch": 16.420926517571885, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20559 + }, + { + "epoch": 16.421725239616613, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 20560 + }, + { + "epoch": 16.42252396166134, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 20561 + }, + { + "epoch": 16.423322683706072, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 20562 + }, + { + "epoch": 16.4241214057508, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 20563 + }, + { + "epoch": 16.424920127795527, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 20564 + }, + { + "epoch": 16.425718849840255, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 20565 + }, + { + "epoch": 16.426517571884983, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 20566 + }, + { + "epoch": 16.427316293929714, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 20567 + }, + { + "epoch": 16.42811501597444, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 20568 + }, + { + "epoch": 16.42891373801917, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 20569 + }, + { + "epoch": 16.429712460063897, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20570 + }, + { + "epoch": 16.430511182108628, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 20571 + }, + { + "epoch": 16.431309904153355, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 20572 + }, + { + "epoch": 16.432108626198083, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20573 + }, + { + "epoch": 16.43290734824281, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 20574 + }, + { + "epoch": 16.43370607028754, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 20575 + }, + { + "epoch": 16.43450479233227, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 20576 + }, + { + "epoch": 16.435303514376997, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 20577 + }, + { + "epoch": 16.436102236421725, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 20578 + }, + { + "epoch": 16.436900958466452, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 20579 + }, + { + "epoch": 16.437699680511184, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 20580 + }, + { + "epoch": 16.43849840255591, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 20581 + }, + { + "epoch": 16.43929712460064, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 20582 + }, + { + "epoch": 16.440095846645367, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 20583 + }, + { + "epoch": 16.440894568690094, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 20584 + }, + { + "epoch": 16.441693290734825, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 20585 + }, + { + "epoch": 16.442492012779553, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 20586 + }, + { + "epoch": 16.44329073482428, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 20587 + }, + { + "epoch": 16.44408945686901, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 20588 + }, + { + "epoch": 16.44488817891374, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 20589 + }, + { + "epoch": 16.445686900958467, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 20590 + }, + { + "epoch": 16.446485623003195, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 20591 + }, + { + "epoch": 16.447284345047922, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 20592 + }, + { + "epoch": 16.44808306709265, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 20593 + }, + { + "epoch": 16.44888178913738, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 20594 + }, + { + "epoch": 16.44968051118211, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 20595 + }, + { + "epoch": 16.450479233226837, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 20596 + }, + { + "epoch": 16.451277955271564, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 20597 + }, + { + "epoch": 16.452076677316295, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 20598 + }, + { + "epoch": 16.452875399361023, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 20599 + }, + { + "epoch": 16.45367412140575, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 20600 + }, + { + "epoch": 16.45447284345048, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 20601 + }, + { + "epoch": 16.455271565495206, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 20602 + }, + { + "epoch": 16.456070287539937, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 20603 + }, + { + "epoch": 16.456869009584665, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 20604 + }, + { + "epoch": 16.457667731629392, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 20605 + }, + { + "epoch": 16.45846645367412, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 20606 + }, + { + "epoch": 16.45926517571885, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 20607 + }, + { + "epoch": 16.46006389776358, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 20608 + }, + { + "epoch": 16.460862619808307, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 20609 + }, + { + "epoch": 16.461661341853034, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 20610 + }, + { + "epoch": 16.462460063897762, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 20611 + }, + { + "epoch": 16.463258785942493, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20612 + }, + { + "epoch": 16.46405750798722, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 20613 + }, + { + "epoch": 16.46485623003195, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 20614 + }, + { + "epoch": 16.465654952076676, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 20615 + }, + { + "epoch": 16.466453674121407, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20616 + }, + { + "epoch": 16.467252396166135, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 20617 + }, + { + "epoch": 16.468051118210862, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 20618 + }, + { + "epoch": 16.46884984025559, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 20619 + }, + { + "epoch": 16.46964856230032, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 20620 + }, + { + "epoch": 16.47044728434505, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 20621 + }, + { + "epoch": 16.471246006389777, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 20622 + }, + { + "epoch": 16.472044728434504, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 20623 + }, + { + "epoch": 16.472843450479232, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 20624 + }, + { + "epoch": 16.473642172523963, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 20625 + }, + { + "epoch": 16.47444089456869, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 20626 + }, + { + "epoch": 16.47523961661342, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 20627 + }, + { + "epoch": 16.476038338658146, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 20628 + }, + { + "epoch": 16.476837060702877, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 20629 + }, + { + "epoch": 16.477635782747605, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 20630 + }, + { + "epoch": 16.478434504792332, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 20631 + }, + { + "epoch": 16.47923322683706, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 20632 + }, + { + "epoch": 16.480031948881788, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 20633 + }, + { + "epoch": 16.48083067092652, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 20634 + }, + { + "epoch": 16.481629392971247, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 20635 + }, + { + "epoch": 16.482428115015974, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 20636 + }, + { + "epoch": 16.483226837060702, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 20637 + }, + { + "epoch": 16.484025559105433, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 20638 + }, + { + "epoch": 16.48482428115016, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 20639 + }, + { + "epoch": 16.48562300319489, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 20640 + }, + { + "epoch": 16.486421725239616, + "grad_norm": 0.365234375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 20641 + }, + { + "epoch": 16.487220447284344, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 20642 + }, + { + "epoch": 16.488019169329075, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 20643 + }, + { + "epoch": 16.488817891373802, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 20644 + }, + { + "epoch": 16.48961661341853, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 20645 + }, + { + "epoch": 16.490415335463258, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 20646 + }, + { + "epoch": 16.49121405750799, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 20647 + }, + { + "epoch": 16.492012779552716, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 20648 + }, + { + "epoch": 16.492811501597444, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 20649 + }, + { + "epoch": 16.49361022364217, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 20650 + }, + { + "epoch": 16.4944089456869, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 20651 + }, + { + "epoch": 16.49520766773163, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20652 + }, + { + "epoch": 16.49600638977636, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 20653 + }, + { + "epoch": 16.496805111821086, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 20654 + }, + { + "epoch": 16.497603833865814, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 20655 + }, + { + "epoch": 16.498402555910545, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 20656 + }, + { + "epoch": 16.499201277955272, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 20657 + }, + { + "epoch": 16.5, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 20658 + }, + { + "epoch": 16.500798722044728, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 20659 + }, + { + "epoch": 16.501597444089455, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 20660 + }, + { + "epoch": 16.502396166134186, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 20661 + }, + { + "epoch": 16.503194888178914, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 20662 + }, + { + "epoch": 16.50399361022364, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 20663 + }, + { + "epoch": 16.50479233226837, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 20664 + }, + { + "epoch": 16.5055910543131, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 20665 + }, + { + "epoch": 16.50638977635783, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20666 + }, + { + "epoch": 16.507188498402556, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 20667 + }, + { + "epoch": 16.507987220447284, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 20668 + }, + { + "epoch": 16.50878594249201, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20669 + }, + { + "epoch": 16.509584664536742, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 20670 + }, + { + "epoch": 16.51038338658147, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 20671 + }, + { + "epoch": 16.511182108626198, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20672 + }, + { + "epoch": 16.511980830670925, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 20673 + }, + { + "epoch": 16.512779552715656, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 20674 + }, + { + "epoch": 16.513578274760384, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 20675 + }, + { + "epoch": 16.51437699680511, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 20676 + }, + { + "epoch": 16.51517571884984, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 20677 + }, + { + "epoch": 16.515974440894567, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 20678 + }, + { + "epoch": 16.516773162939298, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 20679 + }, + { + "epoch": 16.517571884984026, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20680 + }, + { + "epoch": 16.518370607028753, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 20681 + }, + { + "epoch": 16.51916932907348, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 20682 + }, + { + "epoch": 16.519968051118212, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 20683 + }, + { + "epoch": 16.52076677316294, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 20684 + }, + { + "epoch": 16.521565495207668, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 20685 + }, + { + "epoch": 16.522364217252395, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 20686 + }, + { + "epoch": 16.523162939297123, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 20687 + }, + { + "epoch": 16.523961661341854, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 20688 + }, + { + "epoch": 16.52476038338658, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 20689 + }, + { + "epoch": 16.52555910543131, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 20690 + }, + { + "epoch": 16.526357827476037, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 20691 + }, + { + "epoch": 16.527156549520768, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1104, + "step": 20692 + }, + { + "epoch": 16.527955271565496, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 20693 + }, + { + "epoch": 16.528753993610223, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 20694 + }, + { + "epoch": 16.52955271565495, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 20695 + }, + { + "epoch": 16.53035143769968, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 20696 + }, + { + "epoch": 16.53115015974441, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 20697 + }, + { + "epoch": 16.531948881789138, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 20698 + }, + { + "epoch": 16.532747603833865, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 20699 + }, + { + "epoch": 16.533546325878593, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 20700 + }, + { + "epoch": 16.534345047923324, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 20701 + }, + { + "epoch": 16.53514376996805, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 20702 + }, + { + "epoch": 16.53594249201278, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 20703 + }, + { + "epoch": 16.536741214057507, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 20704 + }, + { + "epoch": 16.537539936102235, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20705 + }, + { + "epoch": 16.538338658146966, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 20706 + }, + { + "epoch": 16.539137380191693, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 20707 + }, + { + "epoch": 16.53993610223642, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 20708 + }, + { + "epoch": 16.54073482428115, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 20709 + }, + { + "epoch": 16.54153354632588, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 20710 + }, + { + "epoch": 16.542332268370608, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 20711 + }, + { + "epoch": 16.543130990415335, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 20712 + }, + { + "epoch": 16.543929712460063, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20713 + }, + { + "epoch": 16.544728434504794, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 20714 + }, + { + "epoch": 16.54552715654952, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 20715 + }, + { + "epoch": 16.54632587859425, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 20716 + }, + { + "epoch": 16.547124600638977, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 20717 + }, + { + "epoch": 16.547923322683705, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 20718 + }, + { + "epoch": 16.548722044728436, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 20719 + }, + { + "epoch": 16.549520766773163, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 20720 + }, + { + "epoch": 16.55031948881789, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 20721 + }, + { + "epoch": 16.55111821086262, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 20722 + }, + { + "epoch": 16.55191693290735, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 20723 + }, + { + "epoch": 16.552715654952078, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 20724 + }, + { + "epoch": 16.553514376996805, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 20725 + }, + { + "epoch": 16.554313099041533, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 20726 + }, + { + "epoch": 16.55511182108626, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 20727 + }, + { + "epoch": 16.55591054313099, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 20728 + }, + { + "epoch": 16.55670926517572, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 20729 + }, + { + "epoch": 16.557507987220447, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 20730 + }, + { + "epoch": 16.558306709265175, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 20731 + }, + { + "epoch": 16.559105431309906, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 20732 + }, + { + "epoch": 16.559904153354633, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 20733 + }, + { + "epoch": 16.56070287539936, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 20734 + }, + { + "epoch": 16.56150159744409, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 20735 + }, + { + "epoch": 16.562300319488816, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20736 + }, + { + "epoch": 16.563099041533548, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 20737 + }, + { + "epoch": 16.563897763578275, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 20738 + }, + { + "epoch": 16.564696485623003, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 20739 + }, + { + "epoch": 16.56549520766773, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 20740 + }, + { + "epoch": 16.56629392971246, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 20741 + }, + { + "epoch": 16.56709265175719, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 20742 + }, + { + "epoch": 16.567891373801917, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 20743 + }, + { + "epoch": 16.568690095846645, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 20744 + }, + { + "epoch": 16.569488817891372, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 20745 + }, + { + "epoch": 16.570287539936103, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 20746 + }, + { + "epoch": 16.57108626198083, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 20747 + }, + { + "epoch": 16.57188498402556, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 20748 + }, + { + "epoch": 16.572683706070286, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 20749 + }, + { + "epoch": 16.573482428115017, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 20750 + }, + { + "epoch": 16.574281150159745, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 20751 + }, + { + "epoch": 16.575079872204473, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 20752 + }, + { + "epoch": 16.5758785942492, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 20753 + }, + { + "epoch": 16.576677316293928, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 20754 + }, + { + "epoch": 16.57747603833866, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 20755 + }, + { + "epoch": 16.578274760383387, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 20756 + }, + { + "epoch": 16.579073482428115, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 20757 + }, + { + "epoch": 16.579872204472842, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20758 + }, + { + "epoch": 16.580670926517573, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 20759 + }, + { + "epoch": 16.5814696485623, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 20760 + }, + { + "epoch": 16.58226837060703, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 20761 + }, + { + "epoch": 16.583067092651756, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 20762 + }, + { + "epoch": 16.583865814696484, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20763 + }, + { + "epoch": 16.584664536741215, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 20764 + }, + { + "epoch": 16.585463258785943, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 20765 + }, + { + "epoch": 16.58626198083067, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 20766 + }, + { + "epoch": 16.587060702875398, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 20767 + }, + { + "epoch": 16.58785942492013, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 20768 + }, + { + "epoch": 16.588658146964857, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 20769 + }, + { + "epoch": 16.589456869009584, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 20770 + }, + { + "epoch": 16.590255591054312, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 20771 + }, + { + "epoch": 16.591054313099043, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20772 + }, + { + "epoch": 16.59185303514377, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 20773 + }, + { + "epoch": 16.5926517571885, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 20774 + }, + { + "epoch": 16.593450479233226, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 20775 + }, + { + "epoch": 16.594249201277954, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 20776 + }, + { + "epoch": 16.595047923322685, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 20777 + }, + { + "epoch": 16.595846645367413, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 20778 + }, + { + "epoch": 16.59664536741214, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 20779 + }, + { + "epoch": 16.597444089456868, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 20780 + }, + { + "epoch": 16.5982428115016, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 20781 + }, + { + "epoch": 16.599041533546327, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 20782 + }, + { + "epoch": 16.599840255591054, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 20783 + }, + { + "epoch": 16.600638977635782, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 20784 + }, + { + "epoch": 16.60143769968051, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 20785 + }, + { + "epoch": 16.60223642172524, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 20786 + }, + { + "epoch": 16.60303514376997, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 20787 + }, + { + "epoch": 16.603833865814696, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20788 + }, + { + "epoch": 16.604632587859424, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 20789 + }, + { + "epoch": 16.605431309904155, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 20790 + }, + { + "epoch": 16.606230031948883, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 20791 + }, + { + "epoch": 16.60702875399361, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 20792 + }, + { + "epoch": 16.607827476038338, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 20793 + }, + { + "epoch": 16.608626198083066, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20794 + }, + { + "epoch": 16.609424920127797, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 20795 + }, + { + "epoch": 16.610223642172524, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 20796 + }, + { + "epoch": 16.611022364217252, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 20797 + }, + { + "epoch": 16.61182108626198, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 20798 + }, + { + "epoch": 16.61261980830671, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 20799 + }, + { + "epoch": 16.61341853035144, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 20800 + }, + { + "epoch": 16.614217252396166, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 20801 + }, + { + "epoch": 16.615015974440894, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 20802 + }, + { + "epoch": 16.61581469648562, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 20803 + }, + { + "epoch": 16.616613418530353, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 20804 + }, + { + "epoch": 16.61741214057508, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 20805 + }, + { + "epoch": 16.618210862619808, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 20806 + }, + { + "epoch": 16.619009584664536, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 20807 + }, + { + "epoch": 16.619808306709267, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 20808 + }, + { + "epoch": 16.620607028753994, + "grad_norm": 0.09619140625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 20809 + }, + { + "epoch": 16.621405750798722, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 20810 + }, + { + "epoch": 16.62220447284345, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 20811 + }, + { + "epoch": 16.623003194888177, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 20812 + }, + { + "epoch": 16.62380191693291, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 20813 + }, + { + "epoch": 16.624600638977636, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 20814 + }, + { + "epoch": 16.625399361022364, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 20815 + }, + { + "epoch": 16.62619808306709, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 20816 + }, + { + "epoch": 16.626996805111823, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 20817 + }, + { + "epoch": 16.62779552715655, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 20818 + }, + { + "epoch": 16.628594249201278, + "grad_norm": 0.08837890625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 20819 + }, + { + "epoch": 16.629392971246006, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 20820 + }, + { + "epoch": 16.630191693290733, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 20821 + }, + { + "epoch": 16.630990415335464, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 20822 + }, + { + "epoch": 16.631789137380192, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 20823 + }, + { + "epoch": 16.63258785942492, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 20824 + }, + { + "epoch": 16.633386581469647, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 20825 + }, + { + "epoch": 16.63418530351438, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 20826 + }, + { + "epoch": 16.634984025559106, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 20827 + }, + { + "epoch": 16.635782747603834, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 20828 + }, + { + "epoch": 16.63658146964856, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 20829 + }, + { + "epoch": 16.63738019169329, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 20830 + }, + { + "epoch": 16.63817891373802, + "grad_norm": 0.09521484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 20831 + }, + { + "epoch": 16.638977635782748, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 20832 + }, + { + "epoch": 16.639776357827476, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 20833 + }, + { + "epoch": 16.640575079872203, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20834 + }, + { + "epoch": 16.641373801916934, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20835 + }, + { + "epoch": 16.642172523961662, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 20836 + }, + { + "epoch": 16.64297124600639, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 20837 + }, + { + "epoch": 16.643769968051117, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 20838 + }, + { + "epoch": 16.644568690095845, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 20839 + }, + { + "epoch": 16.645367412140576, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 20840 + }, + { + "epoch": 16.646166134185304, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 20841 + }, + { + "epoch": 16.64696485623003, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 20842 + }, + { + "epoch": 16.64776357827476, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20843 + }, + { + "epoch": 16.64856230031949, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 20844 + }, + { + "epoch": 16.649361022364218, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 20845 + }, + { + "epoch": 16.650159744408946, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 20846 + }, + { + "epoch": 16.650958466453673, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 20847 + }, + { + "epoch": 16.6517571884984, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 20848 + }, + { + "epoch": 16.652555910543132, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 20849 + }, + { + "epoch": 16.65335463258786, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 20850 + }, + { + "epoch": 16.654153354632587, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20851 + }, + { + "epoch": 16.654952076677315, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 20852 + }, + { + "epoch": 16.655750798722046, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 20853 + }, + { + "epoch": 16.656549520766774, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 20854 + }, + { + "epoch": 16.6573482428115, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 20855 + }, + { + "epoch": 16.65814696485623, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 20856 + }, + { + "epoch": 16.658945686900957, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 20857 + }, + { + "epoch": 16.659744408945688, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 20858 + }, + { + "epoch": 16.660543130990416, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 20859 + }, + { + "epoch": 16.661341853035143, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 20860 + }, + { + "epoch": 16.66214057507987, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 20861 + }, + { + "epoch": 16.662939297124602, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 20862 + }, + { + "epoch": 16.66373801916933, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 20863 + }, + { + "epoch": 16.664536741214057, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 20864 + }, + { + "epoch": 16.665335463258785, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 20865 + }, + { + "epoch": 16.666134185303516, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 20866 + }, + { + "epoch": 16.666932907348244, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 20867 + }, + { + "epoch": 16.66773162939297, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 20868 + }, + { + "epoch": 16.6685303514377, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20869 + }, + { + "epoch": 16.669329073482427, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 20870 + }, + { + "epoch": 16.670127795527158, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20871 + }, + { + "epoch": 16.670926517571885, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 20872 + }, + { + "epoch": 16.671725239616613, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 20873 + }, + { + "epoch": 16.67252396166134, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20874 + }, + { + "epoch": 16.673322683706072, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 20875 + }, + { + "epoch": 16.6741214057508, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 20876 + }, + { + "epoch": 16.674920127795527, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 20877 + }, + { + "epoch": 16.675718849840255, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 20878 + }, + { + "epoch": 16.676517571884983, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 20879 + }, + { + "epoch": 16.677316293929714, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 20880 + }, + { + "epoch": 16.67811501597444, + "grad_norm": 0.0966796875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 20881 + }, + { + "epoch": 16.67891373801917, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 20882 + }, + { + "epoch": 16.679712460063897, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 20883 + }, + { + "epoch": 16.680511182108628, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 20884 + }, + { + "epoch": 16.681309904153355, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 20885 + }, + { + "epoch": 16.682108626198083, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 20886 + }, + { + "epoch": 16.68290734824281, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 20887 + }, + { + "epoch": 16.68370607028754, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 20888 + }, + { + "epoch": 16.68450479233227, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 20889 + }, + { + "epoch": 16.685303514376997, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 20890 + }, + { + "epoch": 16.686102236421725, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 20891 + }, + { + "epoch": 16.686900958466452, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 20892 + }, + { + "epoch": 16.687699680511184, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20893 + }, + { + "epoch": 16.68849840255591, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 20894 + }, + { + "epoch": 16.68929712460064, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 20895 + }, + { + "epoch": 16.690095846645367, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20896 + }, + { + "epoch": 16.690894568690094, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 20897 + }, + { + "epoch": 16.691693290734825, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 20898 + }, + { + "epoch": 16.692492012779553, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 20899 + }, + { + "epoch": 16.69329073482428, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 20900 + }, + { + "epoch": 16.69408945686901, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 20901 + }, + { + "epoch": 16.69488817891374, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 20902 + }, + { + "epoch": 16.695686900958467, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 20903 + }, + { + "epoch": 16.696485623003195, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 20904 + }, + { + "epoch": 16.697284345047922, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 20905 + }, + { + "epoch": 16.69808306709265, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 20906 + }, + { + "epoch": 16.69888178913738, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 20907 + }, + { + "epoch": 16.69968051118211, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 20908 + }, + { + "epoch": 16.700479233226837, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 20909 + }, + { + "epoch": 16.701277955271564, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 20910 + }, + { + "epoch": 16.702076677316295, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 20911 + }, + { + "epoch": 16.702875399361023, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 20912 + }, + { + "epoch": 16.70367412140575, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 20913 + }, + { + "epoch": 16.70447284345048, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 20914 + }, + { + "epoch": 16.705271565495206, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20915 + }, + { + "epoch": 16.706070287539937, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 20916 + }, + { + "epoch": 16.706869009584665, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 20917 + }, + { + "epoch": 16.707667731629392, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 20918 + }, + { + "epoch": 16.70846645367412, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 20919 + }, + { + "epoch": 16.70926517571885, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 20920 + }, + { + "epoch": 16.71006389776358, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 20921 + }, + { + "epoch": 16.710862619808307, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 20922 + }, + { + "epoch": 16.711661341853034, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 20923 + }, + { + "epoch": 16.712460063897765, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 20924 + }, + { + "epoch": 16.713258785942493, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 20925 + }, + { + "epoch": 16.71405750798722, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 20926 + }, + { + "epoch": 16.71485623003195, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 20927 + }, + { + "epoch": 16.715654952076676, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 20928 + }, + { + "epoch": 16.716453674121407, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 20929 + }, + { + "epoch": 16.717252396166135, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 20930 + }, + { + "epoch": 16.718051118210862, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 20931 + }, + { + "epoch": 16.71884984025559, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 20932 + }, + { + "epoch": 16.71964856230032, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1122, + "step": 20933 + }, + { + "epoch": 16.72044728434505, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 20934 + }, + { + "epoch": 16.721246006389777, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 20935 + }, + { + "epoch": 16.722044728434504, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 20936 + }, + { + "epoch": 16.722843450479232, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 20937 + }, + { + "epoch": 16.723642172523963, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 20938 + }, + { + "epoch": 16.72444089456869, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 20939 + }, + { + "epoch": 16.72523961661342, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 20940 + }, + { + "epoch": 16.726038338658146, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 20941 + }, + { + "epoch": 16.726837060702877, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 20942 + }, + { + "epoch": 16.727635782747605, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20943 + }, + { + "epoch": 16.728434504792332, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 20944 + }, + { + "epoch": 16.72923322683706, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 20945 + }, + { + "epoch": 16.730031948881788, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20946 + }, + { + "epoch": 16.73083067092652, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 20947 + }, + { + "epoch": 16.731629392971247, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 20948 + }, + { + "epoch": 16.732428115015974, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 20949 + }, + { + "epoch": 16.733226837060702, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 20950 + }, + { + "epoch": 16.734025559105433, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 20951 + }, + { + "epoch": 16.73482428115016, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 20952 + }, + { + "epoch": 16.73562300319489, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20953 + }, + { + "epoch": 16.736421725239616, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 20954 + }, + { + "epoch": 16.737220447284344, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 20955 + }, + { + "epoch": 16.738019169329075, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 20956 + }, + { + "epoch": 16.738817891373802, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 20957 + }, + { + "epoch": 16.73961661341853, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 20958 + }, + { + "epoch": 16.740415335463258, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 20959 + }, + { + "epoch": 16.74121405750799, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 20960 + }, + { + "epoch": 16.742012779552716, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 20961 + }, + { + "epoch": 16.742811501597444, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 20962 + }, + { + "epoch": 16.74361022364217, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 20963 + }, + { + "epoch": 16.7444089456869, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 20964 + }, + { + "epoch": 16.74520766773163, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 20965 + }, + { + "epoch": 16.74600638977636, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 20966 + }, + { + "epoch": 16.746805111821086, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 20967 + }, + { + "epoch": 16.747603833865814, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20968 + }, + { + "epoch": 16.748402555910545, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 20969 + }, + { + "epoch": 16.749201277955272, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 20970 + }, + { + "epoch": 16.75, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 20971 + }, + { + "epoch": 16.750798722044728, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 20972 + }, + { + "epoch": 16.751597444089455, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 20973 + }, + { + "epoch": 16.752396166134186, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 20974 + }, + { + "epoch": 16.753194888178914, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 20975 + }, + { + "epoch": 16.75399361022364, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 20976 + }, + { + "epoch": 16.75479233226837, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 20977 + }, + { + "epoch": 16.7555910543131, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 20978 + }, + { + "epoch": 16.75638977635783, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 20979 + }, + { + "epoch": 16.757188498402556, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 20980 + }, + { + "epoch": 16.757987220447284, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 20981 + }, + { + "epoch": 16.75878594249201, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 20982 + }, + { + "epoch": 16.759584664536742, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 20983 + }, + { + "epoch": 16.76038338658147, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 20984 + }, + { + "epoch": 16.761182108626198, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 20985 + }, + { + "epoch": 16.761980830670925, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 20986 + }, + { + "epoch": 16.762779552715656, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 20987 + }, + { + "epoch": 16.763578274760384, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 20988 + }, + { + "epoch": 16.76437699680511, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 20989 + }, + { + "epoch": 16.76517571884984, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 20990 + }, + { + "epoch": 16.765974440894567, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 20991 + }, + { + "epoch": 16.766773162939298, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 20992 + }, + { + "epoch": 16.767571884984026, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 20993 + }, + { + "epoch": 16.768370607028753, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 20994 + }, + { + "epoch": 16.76916932907348, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 20995 + }, + { + "epoch": 16.769968051118212, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 20996 + }, + { + "epoch": 16.77076677316294, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 20997 + }, + { + "epoch": 16.771565495207668, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 20998 + }, + { + "epoch": 16.772364217252395, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 20999 + }, + { + "epoch": 16.773162939297123, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 21000 + }, + { + "epoch": 16.773961661341854, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 21001 + }, + { + "epoch": 16.77476038338658, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 21002 + }, + { + "epoch": 16.77555910543131, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 21003 + }, + { + "epoch": 16.776357827476037, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 21004 + }, + { + "epoch": 16.777156549520768, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 21005 + }, + { + "epoch": 16.777955271565496, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 21006 + }, + { + "epoch": 16.778753993610223, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 21007 + }, + { + "epoch": 16.77955271565495, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 21008 + }, + { + "epoch": 16.78035143769968, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 21009 + }, + { + "epoch": 16.78115015974441, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 21010 + }, + { + "epoch": 16.781948881789138, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 21011 + }, + { + "epoch": 16.782747603833865, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 21012 + }, + { + "epoch": 16.783546325878593, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 21013 + }, + { + "epoch": 16.784345047923324, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 21014 + }, + { + "epoch": 16.78514376996805, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 21015 + }, + { + "epoch": 16.78594249201278, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 21016 + }, + { + "epoch": 16.786741214057507, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 21017 + }, + { + "epoch": 16.787539936102235, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 21018 + }, + { + "epoch": 16.788338658146966, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 21019 + }, + { + "epoch": 16.789137380191693, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 21020 + }, + { + "epoch": 16.78993610223642, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21021 + }, + { + "epoch": 16.79073482428115, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 21022 + }, + { + "epoch": 16.79153354632588, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 21023 + }, + { + "epoch": 16.792332268370608, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 21024 + }, + { + "epoch": 16.793130990415335, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 21025 + }, + { + "epoch": 16.793929712460063, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 21026 + }, + { + "epoch": 16.794728434504794, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 21027 + }, + { + "epoch": 16.79552715654952, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 21028 + }, + { + "epoch": 16.79632587859425, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21029 + }, + { + "epoch": 16.797124600638977, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 21030 + }, + { + "epoch": 16.797923322683705, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 21031 + }, + { + "epoch": 16.798722044728436, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 21032 + }, + { + "epoch": 16.799520766773163, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 21033 + }, + { + "epoch": 16.80031948881789, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 21034 + }, + { + "epoch": 16.80111821086262, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 21035 + }, + { + "epoch": 16.80191693290735, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 21036 + }, + { + "epoch": 16.802715654952078, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 21037 + }, + { + "epoch": 16.803514376996805, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 21038 + }, + { + "epoch": 16.804313099041533, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 21039 + }, + { + "epoch": 16.80511182108626, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 21040 + }, + { + "epoch": 16.80591054313099, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 21041 + }, + { + "epoch": 16.80670926517572, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 21042 + }, + { + "epoch": 16.807507987220447, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 21043 + }, + { + "epoch": 16.808306709265175, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 21044 + }, + { + "epoch": 16.809105431309906, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 21045 + }, + { + "epoch": 16.809904153354633, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 21046 + }, + { + "epoch": 16.81070287539936, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 21047 + }, + { + "epoch": 16.81150159744409, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 21048 + }, + { + "epoch": 16.812300319488816, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 21049 + }, + { + "epoch": 16.813099041533548, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 21050 + }, + { + "epoch": 16.813897763578275, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 21051 + }, + { + "epoch": 16.814696485623003, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 21052 + }, + { + "epoch": 16.81549520766773, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 21053 + }, + { + "epoch": 16.81629392971246, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 21054 + }, + { + "epoch": 16.81709265175719, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 21055 + }, + { + "epoch": 16.817891373801917, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 21056 + }, + { + "epoch": 16.818690095846645, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 21057 + }, + { + "epoch": 16.819488817891372, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 21058 + }, + { + "epoch": 16.820287539936103, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 21059 + }, + { + "epoch": 16.82108626198083, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 21060 + }, + { + "epoch": 16.82188498402556, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 21061 + }, + { + "epoch": 16.822683706070286, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 21062 + }, + { + "epoch": 16.823482428115017, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 21063 + }, + { + "epoch": 16.824281150159745, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 21064 + }, + { + "epoch": 16.825079872204473, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 21065 + }, + { + "epoch": 16.8258785942492, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 21066 + }, + { + "epoch": 16.826677316293928, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 21067 + }, + { + "epoch": 16.82747603833866, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 21068 + }, + { + "epoch": 16.828274760383387, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 21069 + }, + { + "epoch": 16.829073482428115, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 21070 + }, + { + "epoch": 16.829872204472842, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 21071 + }, + { + "epoch": 16.830670926517573, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 21072 + }, + { + "epoch": 16.8314696485623, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 21073 + }, + { + "epoch": 16.83226837060703, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 21074 + }, + { + "epoch": 16.833067092651756, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 21075 + }, + { + "epoch": 16.833865814696484, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 21076 + }, + { + "epoch": 16.834664536741215, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21077 + }, + { + "epoch": 16.835463258785943, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 21078 + }, + { + "epoch": 16.83626198083067, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 21079 + }, + { + "epoch": 16.837060702875398, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 21080 + }, + { + "epoch": 16.83785942492013, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 21081 + }, + { + "epoch": 16.838658146964857, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 21082 + }, + { + "epoch": 16.839456869009584, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 21083 + }, + { + "epoch": 16.840255591054312, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 21084 + }, + { + "epoch": 16.841054313099043, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 21085 + }, + { + "epoch": 16.84185303514377, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 21086 + }, + { + "epoch": 16.8426517571885, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 21087 + }, + { + "epoch": 16.843450479233226, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 21088 + }, + { + "epoch": 16.844249201277954, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 21089 + }, + { + "epoch": 16.845047923322685, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 21090 + }, + { + "epoch": 16.845846645367413, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 21091 + }, + { + "epoch": 16.84664536741214, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21092 + }, + { + "epoch": 16.847444089456868, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 21093 + }, + { + "epoch": 16.8482428115016, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 21094 + }, + { + "epoch": 16.849041533546327, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 21095 + }, + { + "epoch": 16.849840255591054, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 21096 + }, + { + "epoch": 16.850638977635782, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 21097 + }, + { + "epoch": 16.85143769968051, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 21098 + }, + { + "epoch": 16.85223642172524, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 21099 + }, + { + "epoch": 16.85303514376997, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 21100 + }, + { + "epoch": 16.853833865814696, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21101 + }, + { + "epoch": 16.854632587859424, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21102 + }, + { + "epoch": 16.855431309904155, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 21103 + }, + { + "epoch": 16.856230031948883, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 21104 + }, + { + "epoch": 16.85702875399361, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 21105 + }, + { + "epoch": 16.857827476038338, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 21106 + }, + { + "epoch": 16.858626198083066, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 21107 + }, + { + "epoch": 16.859424920127797, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 21108 + }, + { + "epoch": 16.860223642172524, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 21109 + }, + { + "epoch": 16.861022364217252, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 21110 + }, + { + "epoch": 16.86182108626198, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 21111 + }, + { + "epoch": 16.86261980830671, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 21112 + }, + { + "epoch": 16.86341853035144, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 21113 + }, + { + "epoch": 16.864217252396166, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 21114 + }, + { + "epoch": 16.865015974440894, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 21115 + }, + { + "epoch": 16.86581469648562, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 21116 + }, + { + "epoch": 16.866613418530353, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 21117 + }, + { + "epoch": 16.86741214057508, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 21118 + }, + { + "epoch": 16.868210862619808, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 21119 + }, + { + "epoch": 16.869009584664536, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 21120 + }, + { + "epoch": 16.869808306709267, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 21121 + }, + { + "epoch": 16.870607028753994, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 21122 + }, + { + "epoch": 16.871405750798722, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 21123 + }, + { + "epoch": 16.87220447284345, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 21124 + }, + { + "epoch": 16.873003194888177, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 21125 + }, + { + "epoch": 16.87380191693291, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 21126 + }, + { + "epoch": 16.874600638977636, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 21127 + }, + { + "epoch": 16.875399361022364, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 21128 + }, + { + "epoch": 16.87619808306709, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 21129 + }, + { + "epoch": 16.876996805111823, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 21130 + }, + { + "epoch": 16.87779552715655, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 21131 + }, + { + "epoch": 16.878594249201278, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 21132 + }, + { + "epoch": 16.879392971246006, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 21133 + }, + { + "epoch": 16.880191693290733, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 21134 + }, + { + "epoch": 16.880990415335464, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 21135 + }, + { + "epoch": 16.881789137380192, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 21136 + }, + { + "epoch": 16.88258785942492, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 21137 + }, + { + "epoch": 16.883386581469647, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 21138 + }, + { + "epoch": 16.88418530351438, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 21139 + }, + { + "epoch": 16.884984025559106, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 21140 + }, + { + "epoch": 16.885782747603834, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 21141 + }, + { + "epoch": 16.88658146964856, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21142 + }, + { + "epoch": 16.88738019169329, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 21143 + }, + { + "epoch": 16.88817891373802, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 21144 + }, + { + "epoch": 16.888977635782748, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 21145 + }, + { + "epoch": 16.889776357827476, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 21146 + }, + { + "epoch": 16.890575079872203, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 21147 + }, + { + "epoch": 16.891373801916934, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 21148 + }, + { + "epoch": 16.892172523961662, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 21149 + }, + { + "epoch": 16.89297124600639, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 21150 + }, + { + "epoch": 16.893769968051117, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 21151 + }, + { + "epoch": 16.894568690095845, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21152 + }, + { + "epoch": 16.895367412140576, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 21153 + }, + { + "epoch": 16.896166134185304, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 21154 + }, + { + "epoch": 16.89696485623003, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 21155 + }, + { + "epoch": 16.89776357827476, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 21156 + }, + { + "epoch": 16.89856230031949, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 21157 + }, + { + "epoch": 16.899361022364218, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 21158 + }, + { + "epoch": 16.900159744408946, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 21159 + }, + { + "epoch": 16.900958466453673, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 21160 + }, + { + "epoch": 16.9017571884984, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 21161 + }, + { + "epoch": 16.902555910543132, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21162 + }, + { + "epoch": 16.90335463258786, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 21163 + }, + { + "epoch": 16.904153354632587, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 21164 + }, + { + "epoch": 16.904952076677315, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 21165 + }, + { + "epoch": 16.905750798722046, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 21166 + }, + { + "epoch": 16.906549520766774, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 21167 + }, + { + "epoch": 16.9073482428115, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 21168 + }, + { + "epoch": 16.90814696485623, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.108, + "step": 21169 + }, + { + "epoch": 16.908945686900957, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 21170 + }, + { + "epoch": 16.909744408945688, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 21171 + }, + { + "epoch": 16.910543130990416, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 21172 + }, + { + "epoch": 16.911341853035143, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21173 + }, + { + "epoch": 16.91214057507987, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 21174 + }, + { + "epoch": 16.912939297124602, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 21175 + }, + { + "epoch": 16.91373801916933, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 21176 + }, + { + "epoch": 16.914536741214057, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 21177 + }, + { + "epoch": 16.915335463258785, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 21178 + }, + { + "epoch": 16.916134185303516, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 21179 + }, + { + "epoch": 16.916932907348244, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 21180 + }, + { + "epoch": 16.91773162939297, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 21181 + }, + { + "epoch": 16.9185303514377, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 21182 + }, + { + "epoch": 16.919329073482427, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 21183 + }, + { + "epoch": 16.920127795527158, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21184 + }, + { + "epoch": 16.920926517571885, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 21185 + }, + { + "epoch": 16.921725239616613, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 21186 + }, + { + "epoch": 16.92252396166134, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 21187 + }, + { + "epoch": 16.923322683706072, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 21188 + }, + { + "epoch": 16.9241214057508, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 21189 + }, + { + "epoch": 16.924920127795527, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 21190 + }, + { + "epoch": 16.925718849840255, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 21191 + }, + { + "epoch": 16.926517571884983, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 21192 + }, + { + "epoch": 16.927316293929714, + "grad_norm": 0.09814453125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 21193 + }, + { + "epoch": 16.92811501597444, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 21194 + }, + { + "epoch": 16.92891373801917, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 21195 + }, + { + "epoch": 16.929712460063897, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21196 + }, + { + "epoch": 16.930511182108628, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 21197 + }, + { + "epoch": 16.931309904153355, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 21198 + }, + { + "epoch": 16.932108626198083, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21199 + }, + { + "epoch": 16.93290734824281, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21200 + }, + { + "epoch": 16.93370607028754, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 21201 + }, + { + "epoch": 16.93450479233227, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 21202 + }, + { + "epoch": 16.935303514376997, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 21203 + }, + { + "epoch": 16.936102236421725, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21204 + }, + { + "epoch": 16.936900958466452, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 21205 + }, + { + "epoch": 16.937699680511184, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 21206 + }, + { + "epoch": 16.93849840255591, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 21207 + }, + { + "epoch": 16.93929712460064, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 21208 + }, + { + "epoch": 16.940095846645367, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 21209 + }, + { + "epoch": 16.940894568690094, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 21210 + }, + { + "epoch": 16.941693290734825, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 21211 + }, + { + "epoch": 16.942492012779553, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 21212 + }, + { + "epoch": 16.94329073482428, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 21213 + }, + { + "epoch": 16.94408945686901, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 21214 + }, + { + "epoch": 16.94488817891374, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21215 + }, + { + "epoch": 16.945686900958467, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21216 + }, + { + "epoch": 16.946485623003195, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 21217 + }, + { + "epoch": 16.947284345047922, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 21218 + }, + { + "epoch": 16.94808306709265, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 21219 + }, + { + "epoch": 16.94888178913738, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 21220 + }, + { + "epoch": 16.94968051118211, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 21221 + }, + { + "epoch": 16.950479233226837, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 21222 + }, + { + "epoch": 16.951277955271564, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 21223 + }, + { + "epoch": 16.952076677316295, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 21224 + }, + { + "epoch": 16.952875399361023, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 21225 + }, + { + "epoch": 16.95367412140575, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 21226 + }, + { + "epoch": 16.95447284345048, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 21227 + }, + { + "epoch": 16.955271565495206, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 21228 + }, + { + "epoch": 16.956070287539937, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 21229 + }, + { + "epoch": 16.956869009584665, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 21230 + }, + { + "epoch": 16.957667731629392, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 21231 + }, + { + "epoch": 16.95846645367412, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21232 + }, + { + "epoch": 16.95926517571885, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21233 + }, + { + "epoch": 16.96006389776358, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 21234 + }, + { + "epoch": 16.960862619808307, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 21235 + }, + { + "epoch": 16.961661341853034, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 21236 + }, + { + "epoch": 16.962460063897765, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 21237 + }, + { + "epoch": 16.963258785942493, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 21238 + }, + { + "epoch": 16.96405750798722, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 21239 + }, + { + "epoch": 16.96485623003195, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 21240 + }, + { + "epoch": 16.965654952076676, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 21241 + }, + { + "epoch": 16.966453674121407, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 21242 + }, + { + "epoch": 16.967252396166135, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 21243 + }, + { + "epoch": 16.968051118210862, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 21244 + }, + { + "epoch": 16.96884984025559, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 21245 + }, + { + "epoch": 16.96964856230032, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 21246 + }, + { + "epoch": 16.97044728434505, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 21247 + }, + { + "epoch": 16.971246006389777, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 21248 + }, + { + "epoch": 16.972044728434504, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21249 + }, + { + "epoch": 16.972843450479232, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 21250 + }, + { + "epoch": 16.973642172523963, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 21251 + }, + { + "epoch": 16.97444089456869, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 21252 + }, + { + "epoch": 16.97523961661342, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 21253 + }, + { + "epoch": 16.976038338658146, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 21254 + }, + { + "epoch": 16.976837060702877, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 21255 + }, + { + "epoch": 16.977635782747605, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 21256 + }, + { + "epoch": 16.978434504792332, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 21257 + }, + { + "epoch": 16.97923322683706, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 21258 + }, + { + "epoch": 16.980031948881788, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 21259 + }, + { + "epoch": 16.98083067092652, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 21260 + }, + { + "epoch": 16.981629392971247, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 21261 + }, + { + "epoch": 16.982428115015974, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 21262 + }, + { + "epoch": 16.983226837060702, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 21263 + }, + { + "epoch": 16.984025559105433, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 21264 + }, + { + "epoch": 16.98482428115016, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 21265 + }, + { + "epoch": 16.98562300319489, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 21266 + }, + { + "epoch": 16.986421725239616, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 21267 + }, + { + "epoch": 16.987220447284344, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 21268 + }, + { + "epoch": 16.988019169329075, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 21269 + }, + { + "epoch": 16.988817891373802, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 21270 + }, + { + "epoch": 16.98961661341853, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 21271 + }, + { + "epoch": 16.990415335463258, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 21272 + }, + { + "epoch": 16.99121405750799, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 21273 + }, + { + "epoch": 16.992012779552716, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21274 + }, + { + "epoch": 16.992811501597444, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 21275 + }, + { + "epoch": 16.99361022364217, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 21276 + }, + { + "epoch": 16.9944089456869, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 21277 + }, + { + "epoch": 16.99520766773163, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 21278 + }, + { + "epoch": 16.99600638977636, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 21279 + }, + { + "epoch": 16.996805111821086, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 21280 + }, + { + "epoch": 16.997603833865814, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 21281 + }, + { + "epoch": 16.998402555910545, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 21282 + }, + { + "epoch": 16.999201277955272, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 21283 + }, + { + "epoch": 17.0, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 21284 + }, + { + "epoch": 17.000798722044728, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 21285 + }, + { + "epoch": 17.001597444089455, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 21286 + }, + { + "epoch": 17.002396166134186, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 21287 + }, + { + "epoch": 17.003194888178914, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 21288 + }, + { + "epoch": 17.00399361022364, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 21289 + }, + { + "epoch": 17.00479233226837, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 21290 + }, + { + "epoch": 17.0055910543131, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 21291 + }, + { + "epoch": 17.00638977635783, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21292 + }, + { + "epoch": 17.007188498402556, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 21293 + }, + { + "epoch": 17.007987220447284, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 21294 + }, + { + "epoch": 17.00878594249201, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 21295 + }, + { + "epoch": 17.009584664536742, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21296 + }, + { + "epoch": 17.01038338658147, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 21297 + }, + { + "epoch": 17.011182108626198, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 21298 + }, + { + "epoch": 17.011980830670925, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 21299 + }, + { + "epoch": 17.012779552715656, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 21300 + }, + { + "epoch": 17.013578274760384, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 21301 + }, + { + "epoch": 17.01437699680511, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 21302 + }, + { + "epoch": 17.01517571884984, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 21303 + }, + { + "epoch": 17.015974440894567, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 21304 + }, + { + "epoch": 17.016773162939298, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 21305 + }, + { + "epoch": 17.017571884984026, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 21306 + }, + { + "epoch": 17.018370607028753, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 21307 + }, + { + "epoch": 17.01916932907348, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 21308 + }, + { + "epoch": 17.019968051118212, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21309 + }, + { + "epoch": 17.02076677316294, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 21310 + }, + { + "epoch": 17.021565495207668, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 21311 + }, + { + "epoch": 17.022364217252395, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 21312 + }, + { + "epoch": 17.023162939297123, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 21313 + }, + { + "epoch": 17.023961661341854, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 21314 + }, + { + "epoch": 17.02476038338658, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 21315 + }, + { + "epoch": 17.02555910543131, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 21316 + }, + { + "epoch": 17.026357827476037, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21317 + }, + { + "epoch": 17.027156549520768, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 21318 + }, + { + "epoch": 17.027955271565496, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 21319 + }, + { + "epoch": 17.028753993610223, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 21320 + }, + { + "epoch": 17.02955271565495, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 21321 + }, + { + "epoch": 17.03035143769968, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 21322 + }, + { + "epoch": 17.03115015974441, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 21323 + }, + { + "epoch": 17.031948881789138, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 21324 + }, + { + "epoch": 17.032747603833865, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21325 + }, + { + "epoch": 17.033546325878593, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 21326 + }, + { + "epoch": 17.034345047923324, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 21327 + }, + { + "epoch": 17.03514376996805, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 21328 + }, + { + "epoch": 17.03594249201278, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 21329 + }, + { + "epoch": 17.036741214057507, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 21330 + }, + { + "epoch": 17.037539936102238, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 21331 + }, + { + "epoch": 17.038338658146966, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 21332 + }, + { + "epoch": 17.039137380191693, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 21333 + }, + { + "epoch": 17.03993610223642, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 21334 + }, + { + "epoch": 17.04073482428115, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 21335 + }, + { + "epoch": 17.04153354632588, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 21336 + }, + { + "epoch": 17.042332268370608, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 21337 + }, + { + "epoch": 17.043130990415335, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21338 + }, + { + "epoch": 17.043929712460063, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 21339 + }, + { + "epoch": 17.044728434504794, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 21340 + }, + { + "epoch": 17.04552715654952, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 21341 + }, + { + "epoch": 17.04632587859425, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 21342 + }, + { + "epoch": 17.047124600638977, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 21343 + }, + { + "epoch": 17.047923322683705, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 21344 + }, + { + "epoch": 17.048722044728436, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 21345 + }, + { + "epoch": 17.049520766773163, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 21346 + }, + { + "epoch": 17.05031948881789, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 21347 + }, + { + "epoch": 17.05111821086262, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21348 + }, + { + "epoch": 17.05191693290735, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 21349 + }, + { + "epoch": 17.052715654952078, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 21350 + }, + { + "epoch": 17.053514376996805, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21351 + }, + { + "epoch": 17.054313099041533, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 21352 + }, + { + "epoch": 17.05511182108626, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21353 + }, + { + "epoch": 17.05591054313099, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 21354 + }, + { + "epoch": 17.05670926517572, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 21355 + }, + { + "epoch": 17.057507987220447, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 21356 + }, + { + "epoch": 17.058306709265175, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 21357 + }, + { + "epoch": 17.059105431309906, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 21358 + }, + { + "epoch": 17.059904153354633, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 21359 + }, + { + "epoch": 17.06070287539936, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 21360 + }, + { + "epoch": 17.06150159744409, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 21361 + }, + { + "epoch": 17.062300319488816, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 21362 + }, + { + "epoch": 17.063099041533548, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 21363 + }, + { + "epoch": 17.063897763578275, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 21364 + }, + { + "epoch": 17.064696485623003, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 21365 + }, + { + "epoch": 17.06549520766773, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 21366 + }, + { + "epoch": 17.06629392971246, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 21367 + }, + { + "epoch": 17.06709265175719, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 21368 + }, + { + "epoch": 17.067891373801917, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 21369 + }, + { + "epoch": 17.068690095846645, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 21370 + }, + { + "epoch": 17.069488817891372, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 21371 + }, + { + "epoch": 17.070287539936103, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21372 + }, + { + "epoch": 17.07108626198083, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 21373 + }, + { + "epoch": 17.07188498402556, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 21374 + }, + { + "epoch": 17.072683706070286, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 21375 + }, + { + "epoch": 17.073482428115017, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 21376 + }, + { + "epoch": 17.074281150159745, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 21377 + }, + { + "epoch": 17.075079872204473, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 21378 + }, + { + "epoch": 17.0758785942492, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 21379 + }, + { + "epoch": 17.076677316293928, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 21380 + }, + { + "epoch": 17.07747603833866, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 21381 + }, + { + "epoch": 17.078274760383387, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21382 + }, + { + "epoch": 17.079073482428115, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 21383 + }, + { + "epoch": 17.079872204472842, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 21384 + }, + { + "epoch": 17.080670926517573, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 21385 + }, + { + "epoch": 17.0814696485623, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21386 + }, + { + "epoch": 17.08226837060703, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 21387 + }, + { + "epoch": 17.083067092651756, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 21388 + }, + { + "epoch": 17.083865814696484, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 21389 + }, + { + "epoch": 17.084664536741215, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 21390 + }, + { + "epoch": 17.085463258785943, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 21391 + }, + { + "epoch": 17.08626198083067, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 21392 + }, + { + "epoch": 17.087060702875398, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 21393 + }, + { + "epoch": 17.08785942492013, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 21394 + }, + { + "epoch": 17.088658146964857, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 21395 + }, + { + "epoch": 17.089456869009584, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 21396 + }, + { + "epoch": 17.090255591054312, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 21397 + }, + { + "epoch": 17.09105431309904, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 21398 + }, + { + "epoch": 17.09185303514377, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 21399 + }, + { + "epoch": 17.0926517571885, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 21400 + }, + { + "epoch": 17.093450479233226, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 21401 + }, + { + "epoch": 17.094249201277954, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 21402 + }, + { + "epoch": 17.095047923322685, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 21403 + }, + { + "epoch": 17.095846645367413, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 21404 + }, + { + "epoch": 17.09664536741214, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21405 + }, + { + "epoch": 17.097444089456868, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 21406 + }, + { + "epoch": 17.0982428115016, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 21407 + }, + { + "epoch": 17.099041533546327, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 21408 + }, + { + "epoch": 17.099840255591054, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 21409 + }, + { + "epoch": 17.100638977635782, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 21410 + }, + { + "epoch": 17.10143769968051, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 21411 + }, + { + "epoch": 17.10223642172524, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 21412 + }, + { + "epoch": 17.10303514376997, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 21413 + }, + { + "epoch": 17.103833865814696, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 21414 + }, + { + "epoch": 17.104632587859424, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 21415 + }, + { + "epoch": 17.105431309904155, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 21416 + }, + { + "epoch": 17.106230031948883, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 21417 + }, + { + "epoch": 17.10702875399361, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 21418 + }, + { + "epoch": 17.107827476038338, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 21419 + }, + { + "epoch": 17.108626198083066, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 21420 + }, + { + "epoch": 17.109424920127797, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 21421 + }, + { + "epoch": 17.110223642172524, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 21422 + }, + { + "epoch": 17.111022364217252, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 21423 + }, + { + "epoch": 17.11182108626198, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 21424 + }, + { + "epoch": 17.11261980830671, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 21425 + }, + { + "epoch": 17.11341853035144, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 21426 + }, + { + "epoch": 17.114217252396166, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 21427 + }, + { + "epoch": 17.115015974440894, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 21428 + }, + { + "epoch": 17.11581469648562, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 21429 + }, + { + "epoch": 17.116613418530353, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 21430 + }, + { + "epoch": 17.11741214057508, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 21431 + }, + { + "epoch": 17.118210862619808, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 21432 + }, + { + "epoch": 17.119009584664536, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 21433 + }, + { + "epoch": 17.119808306709267, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 21434 + }, + { + "epoch": 17.120607028753994, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 21435 + }, + { + "epoch": 17.121405750798722, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 21436 + }, + { + "epoch": 17.12220447284345, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21437 + }, + { + "epoch": 17.123003194888177, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 21438 + }, + { + "epoch": 17.12380191693291, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 21439 + }, + { + "epoch": 17.124600638977636, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 21440 + }, + { + "epoch": 17.125399361022364, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 21441 + }, + { + "epoch": 17.12619808306709, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 21442 + }, + { + "epoch": 17.126996805111823, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 21443 + }, + { + "epoch": 17.12779552715655, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 21444 + }, + { + "epoch": 17.128594249201278, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 21445 + }, + { + "epoch": 17.129392971246006, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 21446 + }, + { + "epoch": 17.130191693290733, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 21447 + }, + { + "epoch": 17.130990415335464, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 21448 + }, + { + "epoch": 17.131789137380192, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 21449 + }, + { + "epoch": 17.13258785942492, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 21450 + }, + { + "epoch": 17.133386581469647, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 21451 + }, + { + "epoch": 17.13418530351438, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 21452 + }, + { + "epoch": 17.134984025559106, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 21453 + }, + { + "epoch": 17.135782747603834, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 21454 + }, + { + "epoch": 17.13658146964856, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 21455 + }, + { + "epoch": 17.13738019169329, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 21456 + }, + { + "epoch": 17.13817891373802, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 21457 + }, + { + "epoch": 17.138977635782748, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 21458 + }, + { + "epoch": 17.139776357827476, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 21459 + }, + { + "epoch": 17.140575079872203, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 21460 + }, + { + "epoch": 17.141373801916934, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 21461 + }, + { + "epoch": 17.142172523961662, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 21462 + }, + { + "epoch": 17.14297124600639, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 21463 + }, + { + "epoch": 17.143769968051117, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 21464 + }, + { + "epoch": 17.144568690095845, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 21465 + }, + { + "epoch": 17.145367412140576, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21466 + }, + { + "epoch": 17.146166134185304, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21467 + }, + { + "epoch": 17.14696485623003, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 21468 + }, + { + "epoch": 17.14776357827476, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 21469 + }, + { + "epoch": 17.14856230031949, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21470 + }, + { + "epoch": 17.149361022364218, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 21471 + }, + { + "epoch": 17.150159744408946, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 21472 + }, + { + "epoch": 17.150958466453673, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 21473 + }, + { + "epoch": 17.1517571884984, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 21474 + }, + { + "epoch": 17.152555910543132, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21475 + }, + { + "epoch": 17.15335463258786, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 21476 + }, + { + "epoch": 17.154153354632587, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 21477 + }, + { + "epoch": 17.154952076677315, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 21478 + }, + { + "epoch": 17.155750798722046, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 21479 + }, + { + "epoch": 17.156549520766774, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 21480 + }, + { + "epoch": 17.1573482428115, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21481 + }, + { + "epoch": 17.15814696485623, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 21482 + }, + { + "epoch": 17.15894568690096, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 21483 + }, + { + "epoch": 17.159744408945688, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 21484 + }, + { + "epoch": 17.160543130990416, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 21485 + }, + { + "epoch": 17.161341853035143, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 21486 + }, + { + "epoch": 17.16214057507987, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 21487 + }, + { + "epoch": 17.162939297124602, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 21488 + }, + { + "epoch": 17.16373801916933, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 21489 + }, + { + "epoch": 17.164536741214057, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 21490 + }, + { + "epoch": 17.165335463258785, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 21491 + }, + { + "epoch": 17.166134185303516, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 21492 + }, + { + "epoch": 17.166932907348244, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 21493 + }, + { + "epoch": 17.16773162939297, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 21494 + }, + { + "epoch": 17.1685303514377, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 21495 + }, + { + "epoch": 17.169329073482427, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 21496 + }, + { + "epoch": 17.170127795527158, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 21497 + }, + { + "epoch": 17.170926517571885, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 21498 + }, + { + "epoch": 17.171725239616613, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 21499 + }, + { + "epoch": 17.17252396166134, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 21500 + }, + { + "epoch": 17.173322683706072, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 21501 + }, + { + "epoch": 17.1741214057508, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 21502 + }, + { + "epoch": 17.174920127795527, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 21503 + }, + { + "epoch": 17.175718849840255, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 21504 + }, + { + "epoch": 17.176517571884983, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 21505 + }, + { + "epoch": 17.177316293929714, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 21506 + }, + { + "epoch": 17.17811501597444, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 21507 + }, + { + "epoch": 17.17891373801917, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 21508 + }, + { + "epoch": 17.179712460063897, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 21509 + }, + { + "epoch": 17.180511182108628, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 21510 + }, + { + "epoch": 17.181309904153355, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 21511 + }, + { + "epoch": 17.182108626198083, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 21512 + }, + { + "epoch": 17.18290734824281, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 21513 + }, + { + "epoch": 17.18370607028754, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 21514 + }, + { + "epoch": 17.18450479233227, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 21515 + }, + { + "epoch": 17.185303514376997, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 21516 + }, + { + "epoch": 17.186102236421725, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21517 + }, + { + "epoch": 17.186900958466452, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 21518 + }, + { + "epoch": 17.187699680511184, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21519 + }, + { + "epoch": 17.18849840255591, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 21520 + }, + { + "epoch": 17.18929712460064, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21521 + }, + { + "epoch": 17.190095846645367, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 21522 + }, + { + "epoch": 17.190894568690094, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 21523 + }, + { + "epoch": 17.191693290734825, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 21524 + }, + { + "epoch": 17.192492012779553, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21525 + }, + { + "epoch": 17.19329073482428, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 21526 + }, + { + "epoch": 17.19408945686901, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 21527 + }, + { + "epoch": 17.19488817891374, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 21528 + }, + { + "epoch": 17.195686900958467, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 21529 + }, + { + "epoch": 17.196485623003195, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 21530 + }, + { + "epoch": 17.197284345047922, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 21531 + }, + { + "epoch": 17.19808306709265, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 21532 + }, + { + "epoch": 17.19888178913738, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 21533 + }, + { + "epoch": 17.19968051118211, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 21534 + }, + { + "epoch": 17.200479233226837, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 21535 + }, + { + "epoch": 17.201277955271564, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 21536 + }, + { + "epoch": 17.202076677316295, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 21537 + }, + { + "epoch": 17.202875399361023, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 21538 + }, + { + "epoch": 17.20367412140575, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 21539 + }, + { + "epoch": 17.20447284345048, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 21540 + }, + { + "epoch": 17.205271565495206, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 21541 + }, + { + "epoch": 17.206070287539937, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 21542 + }, + { + "epoch": 17.206869009584665, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 21543 + }, + { + "epoch": 17.207667731629392, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 21544 + }, + { + "epoch": 17.20846645367412, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 21545 + }, + { + "epoch": 17.20926517571885, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 21546 + }, + { + "epoch": 17.21006389776358, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 21547 + }, + { + "epoch": 17.210862619808307, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 21548 + }, + { + "epoch": 17.211661341853034, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 21549 + }, + { + "epoch": 17.212460063897762, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 21550 + }, + { + "epoch": 17.213258785942493, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 21551 + }, + { + "epoch": 17.21405750798722, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 21552 + }, + { + "epoch": 17.21485623003195, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 21553 + }, + { + "epoch": 17.215654952076676, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21554 + }, + { + "epoch": 17.216453674121407, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 21555 + }, + { + "epoch": 17.217252396166135, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 21556 + }, + { + "epoch": 17.218051118210862, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 21557 + }, + { + "epoch": 17.21884984025559, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 21558 + }, + { + "epoch": 17.21964856230032, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 21559 + }, + { + "epoch": 17.22044728434505, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 21560 + }, + { + "epoch": 17.221246006389777, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 21561 + }, + { + "epoch": 17.222044728434504, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 21562 + }, + { + "epoch": 17.222843450479232, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 21563 + }, + { + "epoch": 17.223642172523963, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 21564 + }, + { + "epoch": 17.22444089456869, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 21565 + }, + { + "epoch": 17.22523961661342, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21566 + }, + { + "epoch": 17.226038338658146, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 21567 + }, + { + "epoch": 17.226837060702877, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 21568 + }, + { + "epoch": 17.227635782747605, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21569 + }, + { + "epoch": 17.228434504792332, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 21570 + }, + { + "epoch": 17.22923322683706, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 21571 + }, + { + "epoch": 17.230031948881788, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 21572 + }, + { + "epoch": 17.23083067092652, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 21573 + }, + { + "epoch": 17.231629392971247, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 21574 + }, + { + "epoch": 17.232428115015974, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 21575 + }, + { + "epoch": 17.233226837060702, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 21576 + }, + { + "epoch": 17.234025559105433, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 21577 + }, + { + "epoch": 17.23482428115016, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 21578 + }, + { + "epoch": 17.23562300319489, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 21579 + }, + { + "epoch": 17.236421725239616, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 21580 + }, + { + "epoch": 17.237220447284344, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 21581 + }, + { + "epoch": 17.238019169329075, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 21582 + }, + { + "epoch": 17.238817891373802, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21583 + }, + { + "epoch": 17.23961661341853, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 21584 + }, + { + "epoch": 17.240415335463258, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0751, + "step": 21585 + }, + { + "epoch": 17.24121405750799, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 21586 + }, + { + "epoch": 17.242012779552716, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 21587 + }, + { + "epoch": 17.242811501597444, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 21588 + }, + { + "epoch": 17.24361022364217, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 21589 + }, + { + "epoch": 17.2444089456869, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 21590 + }, + { + "epoch": 17.24520766773163, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 21591 + }, + { + "epoch": 17.24600638977636, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 21592 + }, + { + "epoch": 17.246805111821086, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 21593 + }, + { + "epoch": 17.247603833865814, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 21594 + }, + { + "epoch": 17.248402555910545, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21595 + }, + { + "epoch": 17.249201277955272, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 21596 + }, + { + "epoch": 17.25, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 21597 + }, + { + "epoch": 17.250798722044728, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 21598 + }, + { + "epoch": 17.251597444089455, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 21599 + }, + { + "epoch": 17.252396166134186, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 21600 + }, + { + "epoch": 17.253194888178914, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 21601 + }, + { + "epoch": 17.25399361022364, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21602 + }, + { + "epoch": 17.25479233226837, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 21603 + }, + { + "epoch": 17.2555910543131, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 21604 + }, + { + "epoch": 17.25638977635783, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 21605 + }, + { + "epoch": 17.257188498402556, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21606 + }, + { + "epoch": 17.257987220447284, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 21607 + }, + { + "epoch": 17.25878594249201, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 21608 + }, + { + "epoch": 17.259584664536742, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 21609 + }, + { + "epoch": 17.26038338658147, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21610 + }, + { + "epoch": 17.261182108626198, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 21611 + }, + { + "epoch": 17.261980830670925, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 21612 + }, + { + "epoch": 17.262779552715656, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 21613 + }, + { + "epoch": 17.263578274760384, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 21614 + }, + { + "epoch": 17.26437699680511, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 21615 + }, + { + "epoch": 17.26517571884984, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 21616 + }, + { + "epoch": 17.265974440894567, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 21617 + }, + { + "epoch": 17.266773162939298, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 21618 + }, + { + "epoch": 17.267571884984026, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 21619 + }, + { + "epoch": 17.268370607028753, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 21620 + }, + { + "epoch": 17.26916932907348, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 21621 + }, + { + "epoch": 17.269968051118212, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 21622 + }, + { + "epoch": 17.27076677316294, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 21623 + }, + { + "epoch": 17.271565495207668, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 21624 + }, + { + "epoch": 17.272364217252395, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 21625 + }, + { + "epoch": 17.273162939297123, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 21626 + }, + { + "epoch": 17.273961661341854, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 21627 + }, + { + "epoch": 17.27476038338658, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 21628 + }, + { + "epoch": 17.27555910543131, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 21629 + }, + { + "epoch": 17.276357827476037, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 21630 + }, + { + "epoch": 17.277156549520768, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 21631 + }, + { + "epoch": 17.277955271565496, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 21632 + }, + { + "epoch": 17.278753993610223, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 21633 + }, + { + "epoch": 17.27955271565495, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 21634 + }, + { + "epoch": 17.28035143769968, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 21635 + }, + { + "epoch": 17.28115015974441, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 21636 + }, + { + "epoch": 17.281948881789138, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21637 + }, + { + "epoch": 17.282747603833865, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 21638 + }, + { + "epoch": 17.283546325878593, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 21639 + }, + { + "epoch": 17.284345047923324, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 21640 + }, + { + "epoch": 17.28514376996805, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21641 + }, + { + "epoch": 17.28594249201278, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 21642 + }, + { + "epoch": 17.286741214057507, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 21643 + }, + { + "epoch": 17.287539936102238, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 21644 + }, + { + "epoch": 17.288338658146966, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 21645 + }, + { + "epoch": 17.289137380191693, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 21646 + }, + { + "epoch": 17.28993610223642, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 21647 + }, + { + "epoch": 17.29073482428115, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 21648 + }, + { + "epoch": 17.29153354632588, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 21649 + }, + { + "epoch": 17.292332268370608, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 21650 + }, + { + "epoch": 17.293130990415335, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 21651 + }, + { + "epoch": 17.293929712460063, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 21652 + }, + { + "epoch": 17.294728434504794, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 21653 + }, + { + "epoch": 17.29552715654952, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 21654 + }, + { + "epoch": 17.29632587859425, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 21655 + }, + { + "epoch": 17.297124600638977, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 21656 + }, + { + "epoch": 17.297923322683705, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 21657 + }, + { + "epoch": 17.298722044728436, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 21658 + }, + { + "epoch": 17.299520766773163, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 21659 + }, + { + "epoch": 17.30031948881789, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 21660 + }, + { + "epoch": 17.30111821086262, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 21661 + }, + { + "epoch": 17.30191693290735, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21662 + }, + { + "epoch": 17.302715654952078, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21663 + }, + { + "epoch": 17.303514376996805, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 21664 + }, + { + "epoch": 17.304313099041533, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 21665 + }, + { + "epoch": 17.30511182108626, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 21666 + }, + { + "epoch": 17.30591054313099, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 21667 + }, + { + "epoch": 17.30670926517572, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 21668 + }, + { + "epoch": 17.307507987220447, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 21669 + }, + { + "epoch": 17.308306709265175, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 21670 + }, + { + "epoch": 17.309105431309906, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21671 + }, + { + "epoch": 17.309904153354633, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 21672 + }, + { + "epoch": 17.31070287539936, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 21673 + }, + { + "epoch": 17.31150159744409, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 21674 + }, + { + "epoch": 17.312300319488816, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 21675 + }, + { + "epoch": 17.313099041533548, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 21676 + }, + { + "epoch": 17.313897763578275, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 21677 + }, + { + "epoch": 17.314696485623003, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 21678 + }, + { + "epoch": 17.31549520766773, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 21679 + }, + { + "epoch": 17.31629392971246, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 21680 + }, + { + "epoch": 17.31709265175719, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 21681 + }, + { + "epoch": 17.317891373801917, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 21682 + }, + { + "epoch": 17.318690095846645, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 21683 + }, + { + "epoch": 17.319488817891372, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 21684 + }, + { + "epoch": 17.320287539936103, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 21685 + }, + { + "epoch": 17.32108626198083, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 21686 + }, + { + "epoch": 17.32188498402556, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21687 + }, + { + "epoch": 17.322683706070286, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 21688 + }, + { + "epoch": 17.323482428115017, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 21689 + }, + { + "epoch": 17.324281150159745, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 21690 + }, + { + "epoch": 17.325079872204473, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 21691 + }, + { + "epoch": 17.3258785942492, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 21692 + }, + { + "epoch": 17.326677316293928, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 21693 + }, + { + "epoch": 17.32747603833866, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 21694 + }, + { + "epoch": 17.328274760383387, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 21695 + }, + { + "epoch": 17.329073482428115, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 21696 + }, + { + "epoch": 17.329872204472842, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 21697 + }, + { + "epoch": 17.330670926517573, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 21698 + }, + { + "epoch": 17.3314696485623, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 21699 + }, + { + "epoch": 17.33226837060703, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21700 + }, + { + "epoch": 17.333067092651756, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 21701 + }, + { + "epoch": 17.333865814696484, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 21702 + }, + { + "epoch": 17.334664536741215, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 21703 + }, + { + "epoch": 17.335463258785943, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 21704 + }, + { + "epoch": 17.33626198083067, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 21705 + }, + { + "epoch": 17.337060702875398, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21706 + }, + { + "epoch": 17.33785942492013, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 21707 + }, + { + "epoch": 17.338658146964857, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21708 + }, + { + "epoch": 17.339456869009584, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 21709 + }, + { + "epoch": 17.340255591054312, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 21710 + }, + { + "epoch": 17.341054313099043, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 21711 + }, + { + "epoch": 17.34185303514377, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 21712 + }, + { + "epoch": 17.3426517571885, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 21713 + }, + { + "epoch": 17.343450479233226, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 21714 + }, + { + "epoch": 17.344249201277954, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 21715 + }, + { + "epoch": 17.345047923322685, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 21716 + }, + { + "epoch": 17.345846645367413, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21717 + }, + { + "epoch": 17.34664536741214, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 21718 + }, + { + "epoch": 17.347444089456868, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 21719 + }, + { + "epoch": 17.3482428115016, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 21720 + }, + { + "epoch": 17.349041533546327, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 21721 + }, + { + "epoch": 17.349840255591054, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 21722 + }, + { + "epoch": 17.350638977635782, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21723 + }, + { + "epoch": 17.35143769968051, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 21724 + }, + { + "epoch": 17.35223642172524, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 21725 + }, + { + "epoch": 17.35303514376997, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 21726 + }, + { + "epoch": 17.353833865814696, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 21727 + }, + { + "epoch": 17.354632587859424, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 21728 + }, + { + "epoch": 17.355431309904155, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 21729 + }, + { + "epoch": 17.356230031948883, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 21730 + }, + { + "epoch": 17.35702875399361, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 21731 + }, + { + "epoch": 17.357827476038338, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 21732 + }, + { + "epoch": 17.358626198083066, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 21733 + }, + { + "epoch": 17.359424920127797, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 21734 + }, + { + "epoch": 17.360223642172524, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 21735 + }, + { + "epoch": 17.361022364217252, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 21736 + }, + { + "epoch": 17.36182108626198, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 21737 + }, + { + "epoch": 17.36261980830671, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 21738 + }, + { + "epoch": 17.36341853035144, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 21739 + }, + { + "epoch": 17.364217252396166, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 21740 + }, + { + "epoch": 17.365015974440894, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21741 + }, + { + "epoch": 17.36581469648562, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 21742 + }, + { + "epoch": 17.366613418530353, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 21743 + }, + { + "epoch": 17.36741214057508, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 21744 + }, + { + "epoch": 17.368210862619808, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 21745 + }, + { + "epoch": 17.369009584664536, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 21746 + }, + { + "epoch": 17.369808306709267, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 21747 + }, + { + "epoch": 17.370607028753994, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 21748 + }, + { + "epoch": 17.371405750798722, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 21749 + }, + { + "epoch": 17.37220447284345, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 21750 + }, + { + "epoch": 17.373003194888177, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 21751 + }, + { + "epoch": 17.37380191693291, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 21752 + }, + { + "epoch": 17.374600638977636, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21753 + }, + { + "epoch": 17.375399361022364, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 21754 + }, + { + "epoch": 17.37619808306709, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 21755 + }, + { + "epoch": 17.376996805111823, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 21756 + }, + { + "epoch": 17.37779552715655, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 21757 + }, + { + "epoch": 17.378594249201278, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 21758 + }, + { + "epoch": 17.379392971246006, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 21759 + }, + { + "epoch": 17.380191693290733, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 21760 + }, + { + "epoch": 17.380990415335464, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 21761 + }, + { + "epoch": 17.381789137380192, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 21762 + }, + { + "epoch": 17.38258785942492, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21763 + }, + { + "epoch": 17.383386581469647, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 21764 + }, + { + "epoch": 17.38418530351438, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 21765 + }, + { + "epoch": 17.384984025559106, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 21766 + }, + { + "epoch": 17.385782747603834, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 21767 + }, + { + "epoch": 17.38658146964856, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 21768 + }, + { + "epoch": 17.38738019169329, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 21769 + }, + { + "epoch": 17.38817891373802, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 21770 + }, + { + "epoch": 17.388977635782748, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 21771 + }, + { + "epoch": 17.389776357827476, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 21772 + }, + { + "epoch": 17.390575079872203, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 21773 + }, + { + "epoch": 17.391373801916934, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 21774 + }, + { + "epoch": 17.392172523961662, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 21775 + }, + { + "epoch": 17.39297124600639, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 21776 + }, + { + "epoch": 17.393769968051117, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 21777 + }, + { + "epoch": 17.394568690095845, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 21778 + }, + { + "epoch": 17.395367412140576, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 21779 + }, + { + "epoch": 17.396166134185304, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 21780 + }, + { + "epoch": 17.39696485623003, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 21781 + }, + { + "epoch": 17.39776357827476, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 21782 + }, + { + "epoch": 17.39856230031949, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 21783 + }, + { + "epoch": 17.399361022364218, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 21784 + }, + { + "epoch": 17.400159744408946, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 21785 + }, + { + "epoch": 17.400958466453673, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21786 + }, + { + "epoch": 17.4017571884984, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 21787 + }, + { + "epoch": 17.402555910543132, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 21788 + }, + { + "epoch": 17.40335463258786, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 21789 + }, + { + "epoch": 17.404153354632587, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 21790 + }, + { + "epoch": 17.404952076677315, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 21791 + }, + { + "epoch": 17.405750798722046, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 21792 + }, + { + "epoch": 17.406549520766774, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 21793 + }, + { + "epoch": 17.4073482428115, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 21794 + }, + { + "epoch": 17.40814696485623, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 21795 + }, + { + "epoch": 17.408945686900957, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 21796 + }, + { + "epoch": 17.409744408945688, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 21797 + }, + { + "epoch": 17.410543130990416, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 21798 + }, + { + "epoch": 17.411341853035143, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 21799 + }, + { + "epoch": 17.41214057507987, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 21800 + }, + { + "epoch": 17.412939297124602, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 21801 + }, + { + "epoch": 17.41373801916933, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 21802 + }, + { + "epoch": 17.414536741214057, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 21803 + }, + { + "epoch": 17.415335463258785, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 21804 + }, + { + "epoch": 17.416134185303516, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 21805 + }, + { + "epoch": 17.416932907348244, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 21806 + }, + { + "epoch": 17.41773162939297, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 21807 + }, + { + "epoch": 17.4185303514377, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 21808 + }, + { + "epoch": 17.419329073482427, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 21809 + }, + { + "epoch": 17.420127795527158, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 21810 + }, + { + "epoch": 17.420926517571885, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 21811 + }, + { + "epoch": 17.421725239616613, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 21812 + }, + { + "epoch": 17.42252396166134, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 21813 + }, + { + "epoch": 17.423322683706072, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 21814 + }, + { + "epoch": 17.4241214057508, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 21815 + }, + { + "epoch": 17.424920127795527, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 21816 + }, + { + "epoch": 17.425718849840255, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 21817 + }, + { + "epoch": 17.426517571884983, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 21818 + }, + { + "epoch": 17.427316293929714, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21819 + }, + { + "epoch": 17.42811501597444, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 21820 + }, + { + "epoch": 17.42891373801917, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 21821 + }, + { + "epoch": 17.429712460063897, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 21822 + }, + { + "epoch": 17.430511182108628, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 21823 + }, + { + "epoch": 17.431309904153355, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 21824 + }, + { + "epoch": 17.432108626198083, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 21825 + }, + { + "epoch": 17.43290734824281, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 21826 + }, + { + "epoch": 17.43370607028754, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 21827 + }, + { + "epoch": 17.43450479233227, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21828 + }, + { + "epoch": 17.435303514376997, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 21829 + }, + { + "epoch": 17.436102236421725, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 21830 + }, + { + "epoch": 17.436900958466452, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 21831 + }, + { + "epoch": 17.437699680511184, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21832 + }, + { + "epoch": 17.43849840255591, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 21833 + }, + { + "epoch": 17.43929712460064, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 21834 + }, + { + "epoch": 17.440095846645367, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 21835 + }, + { + "epoch": 17.440894568690094, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 21836 + }, + { + "epoch": 17.441693290734825, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 21837 + }, + { + "epoch": 17.442492012779553, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 21838 + }, + { + "epoch": 17.44329073482428, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 21839 + }, + { + "epoch": 17.44408945686901, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 21840 + }, + { + "epoch": 17.44488817891374, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21841 + }, + { + "epoch": 17.445686900958467, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 21842 + }, + { + "epoch": 17.446485623003195, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 21843 + }, + { + "epoch": 17.447284345047922, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 21844 + }, + { + "epoch": 17.44808306709265, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 21845 + }, + { + "epoch": 17.44888178913738, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 21846 + }, + { + "epoch": 17.44968051118211, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 21847 + }, + { + "epoch": 17.450479233226837, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 21848 + }, + { + "epoch": 17.451277955271564, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 21849 + }, + { + "epoch": 17.452076677316295, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 21850 + }, + { + "epoch": 17.452875399361023, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 21851 + }, + { + "epoch": 17.45367412140575, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 21852 + }, + { + "epoch": 17.45447284345048, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 21853 + }, + { + "epoch": 17.455271565495206, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 21854 + }, + { + "epoch": 17.456070287539937, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 21855 + }, + { + "epoch": 17.456869009584665, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 21856 + }, + { + "epoch": 17.457667731629392, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 21857 + }, + { + "epoch": 17.45846645367412, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 21858 + }, + { + "epoch": 17.45926517571885, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 21859 + }, + { + "epoch": 17.46006389776358, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 21860 + }, + { + "epoch": 17.460862619808307, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 21861 + }, + { + "epoch": 17.461661341853034, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 21862 + }, + { + "epoch": 17.462460063897762, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 21863 + }, + { + "epoch": 17.463258785942493, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 21864 + }, + { + "epoch": 17.46405750798722, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 21865 + }, + { + "epoch": 17.46485623003195, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 21866 + }, + { + "epoch": 17.465654952076676, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 21867 + }, + { + "epoch": 17.466453674121407, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 21868 + }, + { + "epoch": 17.467252396166135, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 21869 + }, + { + "epoch": 17.468051118210862, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 21870 + }, + { + "epoch": 17.46884984025559, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 21871 + }, + { + "epoch": 17.46964856230032, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 21872 + }, + { + "epoch": 17.47044728434505, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21873 + }, + { + "epoch": 17.471246006389777, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 21874 + }, + { + "epoch": 17.472044728434504, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 21875 + }, + { + "epoch": 17.472843450479232, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 21876 + }, + { + "epoch": 17.473642172523963, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 21877 + }, + { + "epoch": 17.47444089456869, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 21878 + }, + { + "epoch": 17.47523961661342, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 21879 + }, + { + "epoch": 17.476038338658146, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 21880 + }, + { + "epoch": 17.476837060702877, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 21881 + }, + { + "epoch": 17.477635782747605, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 21882 + }, + { + "epoch": 17.478434504792332, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 21883 + }, + { + "epoch": 17.47923322683706, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21884 + }, + { + "epoch": 17.480031948881788, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 21885 + }, + { + "epoch": 17.48083067092652, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 21886 + }, + { + "epoch": 17.481629392971247, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 21887 + }, + { + "epoch": 17.482428115015974, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 21888 + }, + { + "epoch": 17.483226837060702, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 21889 + }, + { + "epoch": 17.484025559105433, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 21890 + }, + { + "epoch": 17.48482428115016, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 21891 + }, + { + "epoch": 17.48562300319489, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 21892 + }, + { + "epoch": 17.486421725239616, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 21893 + }, + { + "epoch": 17.487220447284344, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 21894 + }, + { + "epoch": 17.488019169329075, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 21895 + }, + { + "epoch": 17.488817891373802, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 21896 + }, + { + "epoch": 17.48961661341853, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 21897 + }, + { + "epoch": 17.490415335463258, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 21898 + }, + { + "epoch": 17.49121405750799, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 21899 + }, + { + "epoch": 17.492012779552716, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 21900 + }, + { + "epoch": 17.492811501597444, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 21901 + }, + { + "epoch": 17.49361022364217, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 21902 + }, + { + "epoch": 17.4944089456869, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 21903 + }, + { + "epoch": 17.49520766773163, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 21904 + }, + { + "epoch": 17.49600638977636, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21905 + }, + { + "epoch": 17.496805111821086, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 21906 + }, + { + "epoch": 17.497603833865814, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 21907 + }, + { + "epoch": 17.498402555910545, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 21908 + }, + { + "epoch": 17.499201277955272, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 21909 + }, + { + "epoch": 17.5, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 21910 + }, + { + "epoch": 17.500798722044728, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 21911 + }, + { + "epoch": 17.501597444089455, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 21912 + }, + { + "epoch": 17.502396166134186, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 21913 + }, + { + "epoch": 17.503194888178914, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 21914 + }, + { + "epoch": 17.50399361022364, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 21915 + }, + { + "epoch": 17.50479233226837, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 21916 + }, + { + "epoch": 17.5055910543131, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 21917 + }, + { + "epoch": 17.50638977635783, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 21918 + }, + { + "epoch": 17.507188498402556, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 21919 + }, + { + "epoch": 17.507987220447284, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21920 + }, + { + "epoch": 17.50878594249201, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 21921 + }, + { + "epoch": 17.509584664536742, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 21922 + }, + { + "epoch": 17.51038338658147, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 21923 + }, + { + "epoch": 17.511182108626198, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 21924 + }, + { + "epoch": 17.511980830670925, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 21925 + }, + { + "epoch": 17.512779552715656, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 21926 + }, + { + "epoch": 17.513578274760384, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 21927 + }, + { + "epoch": 17.51437699680511, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 21928 + }, + { + "epoch": 17.51517571884984, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 21929 + }, + { + "epoch": 17.515974440894567, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 21930 + }, + { + "epoch": 17.516773162939298, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 21931 + }, + { + "epoch": 17.517571884984026, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21932 + }, + { + "epoch": 17.518370607028753, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 21933 + }, + { + "epoch": 17.51916932907348, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 21934 + }, + { + "epoch": 17.519968051118212, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 21935 + }, + { + "epoch": 17.52076677316294, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 21936 + }, + { + "epoch": 17.521565495207668, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 21937 + }, + { + "epoch": 17.522364217252395, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 21938 + }, + { + "epoch": 17.523162939297123, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 21939 + }, + { + "epoch": 17.523961661341854, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 21940 + }, + { + "epoch": 17.52476038338658, + "grad_norm": 0.3359375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 21941 + }, + { + "epoch": 17.52555910543131, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 21942 + }, + { + "epoch": 17.526357827476037, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 21943 + }, + { + "epoch": 17.527156549520768, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 21944 + }, + { + "epoch": 17.527955271565496, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 21945 + }, + { + "epoch": 17.528753993610223, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 21946 + }, + { + "epoch": 17.52955271565495, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 21947 + }, + { + "epoch": 17.53035143769968, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 21948 + }, + { + "epoch": 17.53115015974441, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 21949 + }, + { + "epoch": 17.531948881789138, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21950 + }, + { + "epoch": 17.532747603833865, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 21951 + }, + { + "epoch": 17.533546325878593, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21952 + }, + { + "epoch": 17.534345047923324, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 21953 + }, + { + "epoch": 17.53514376996805, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 21954 + }, + { + "epoch": 17.53594249201278, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 21955 + }, + { + "epoch": 17.536741214057507, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 21956 + }, + { + "epoch": 17.537539936102235, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 21957 + }, + { + "epoch": 17.538338658146966, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 21958 + }, + { + "epoch": 17.539137380191693, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 21959 + }, + { + "epoch": 17.53993610223642, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 21960 + }, + { + "epoch": 17.54073482428115, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 21961 + }, + { + "epoch": 17.54153354632588, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 21962 + }, + { + "epoch": 17.542332268370608, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 21963 + }, + { + "epoch": 17.543130990415335, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 21964 + }, + { + "epoch": 17.543929712460063, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 21965 + }, + { + "epoch": 17.544728434504794, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21966 + }, + { + "epoch": 17.54552715654952, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 21967 + }, + { + "epoch": 17.54632587859425, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 21968 + }, + { + "epoch": 17.547124600638977, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 21969 + }, + { + "epoch": 17.547923322683705, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 21970 + }, + { + "epoch": 17.548722044728436, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 21971 + }, + { + "epoch": 17.549520766773163, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 21972 + }, + { + "epoch": 17.55031948881789, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 21973 + }, + { + "epoch": 17.55111821086262, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 21974 + }, + { + "epoch": 17.55191693290735, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 21975 + }, + { + "epoch": 17.552715654952078, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 21976 + }, + { + "epoch": 17.553514376996805, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 21977 + }, + { + "epoch": 17.554313099041533, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 21978 + }, + { + "epoch": 17.55511182108626, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 21979 + }, + { + "epoch": 17.55591054313099, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 21980 + }, + { + "epoch": 17.55670926517572, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 21981 + }, + { + "epoch": 17.557507987220447, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 21982 + }, + { + "epoch": 17.558306709265175, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 21983 + }, + { + "epoch": 17.559105431309906, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 21984 + }, + { + "epoch": 17.559904153354633, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 21985 + }, + { + "epoch": 17.56070287539936, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 21986 + }, + { + "epoch": 17.56150159744409, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 21987 + }, + { + "epoch": 17.562300319488816, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 21988 + }, + { + "epoch": 17.563099041533548, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 21989 + }, + { + "epoch": 17.563897763578275, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 21990 + }, + { + "epoch": 17.564696485623003, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 21991 + }, + { + "epoch": 17.56549520766773, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 21992 + }, + { + "epoch": 17.56629392971246, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 21993 + }, + { + "epoch": 17.56709265175719, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 21994 + }, + { + "epoch": 17.567891373801917, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 21995 + }, + { + "epoch": 17.568690095846645, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 21996 + }, + { + "epoch": 17.569488817891372, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 21997 + }, + { + "epoch": 17.570287539936103, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 21998 + }, + { + "epoch": 17.57108626198083, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 21999 + }, + { + "epoch": 17.57188498402556, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1086, + "step": 22000 + }, + { + "epoch": 17.572683706070286, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 22001 + }, + { + "epoch": 17.573482428115017, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 22002 + }, + { + "epoch": 17.574281150159745, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 22003 + }, + { + "epoch": 17.575079872204473, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 22004 + }, + { + "epoch": 17.5758785942492, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 22005 + }, + { + "epoch": 17.576677316293928, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22006 + }, + { + "epoch": 17.57747603833866, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 22007 + }, + { + "epoch": 17.578274760383387, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22008 + }, + { + "epoch": 17.579073482428115, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 22009 + }, + { + "epoch": 17.579872204472842, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 22010 + }, + { + "epoch": 17.580670926517573, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22011 + }, + { + "epoch": 17.5814696485623, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 22012 + }, + { + "epoch": 17.58226837060703, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 22013 + }, + { + "epoch": 17.583067092651756, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22014 + }, + { + "epoch": 17.583865814696484, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 22015 + }, + { + "epoch": 17.584664536741215, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 22016 + }, + { + "epoch": 17.585463258785943, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 22017 + }, + { + "epoch": 17.58626198083067, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 22018 + }, + { + "epoch": 17.587060702875398, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 22019 + }, + { + "epoch": 17.58785942492013, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 22020 + }, + { + "epoch": 17.588658146964857, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 22021 + }, + { + "epoch": 17.589456869009584, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 22022 + }, + { + "epoch": 17.590255591054312, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 22023 + }, + { + "epoch": 17.591054313099043, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 22024 + }, + { + "epoch": 17.59185303514377, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22025 + }, + { + "epoch": 17.5926517571885, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 22026 + }, + { + "epoch": 17.593450479233226, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 22027 + }, + { + "epoch": 17.594249201277954, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 22028 + }, + { + "epoch": 17.595047923322685, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 22029 + }, + { + "epoch": 17.595846645367413, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 22030 + }, + { + "epoch": 17.59664536741214, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 22031 + }, + { + "epoch": 17.597444089456868, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 22032 + }, + { + "epoch": 17.5982428115016, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 22033 + }, + { + "epoch": 17.599041533546327, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 22034 + }, + { + "epoch": 17.599840255591054, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 22035 + }, + { + "epoch": 17.600638977635782, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 22036 + }, + { + "epoch": 17.60143769968051, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 22037 + }, + { + "epoch": 17.60223642172524, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22038 + }, + { + "epoch": 17.60303514376997, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 22039 + }, + { + "epoch": 17.603833865814696, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 22040 + }, + { + "epoch": 17.604632587859424, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 22041 + }, + { + "epoch": 17.605431309904155, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 22042 + }, + { + "epoch": 17.606230031948883, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 22043 + }, + { + "epoch": 17.60702875399361, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 22044 + }, + { + "epoch": 17.607827476038338, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 22045 + }, + { + "epoch": 17.608626198083066, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22046 + }, + { + "epoch": 17.609424920127797, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 22047 + }, + { + "epoch": 17.610223642172524, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 22048 + }, + { + "epoch": 17.611022364217252, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 22049 + }, + { + "epoch": 17.61182108626198, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 22050 + }, + { + "epoch": 17.61261980830671, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 22051 + }, + { + "epoch": 17.61341853035144, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 22052 + }, + { + "epoch": 17.614217252396166, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 22053 + }, + { + "epoch": 17.615015974440894, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 22054 + }, + { + "epoch": 17.61581469648562, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22055 + }, + { + "epoch": 17.616613418530353, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 22056 + }, + { + "epoch": 17.61741214057508, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22057 + }, + { + "epoch": 17.618210862619808, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 22058 + }, + { + "epoch": 17.619009584664536, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 22059 + }, + { + "epoch": 17.619808306709267, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 22060 + }, + { + "epoch": 17.620607028753994, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 22061 + }, + { + "epoch": 17.621405750798722, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 22062 + }, + { + "epoch": 17.62220447284345, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 22063 + }, + { + "epoch": 17.623003194888177, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 22064 + }, + { + "epoch": 17.62380191693291, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 22065 + }, + { + "epoch": 17.624600638977636, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 22066 + }, + { + "epoch": 17.625399361022364, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 22067 + }, + { + "epoch": 17.62619808306709, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 22068 + }, + { + "epoch": 17.626996805111823, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 22069 + }, + { + "epoch": 17.62779552715655, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 22070 + }, + { + "epoch": 17.628594249201278, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 22071 + }, + { + "epoch": 17.629392971246006, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 22072 + }, + { + "epoch": 17.630191693290733, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 22073 + }, + { + "epoch": 17.630990415335464, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 22074 + }, + { + "epoch": 17.631789137380192, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 22075 + }, + { + "epoch": 17.63258785942492, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 22076 + }, + { + "epoch": 17.633386581469647, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 22077 + }, + { + "epoch": 17.63418530351438, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 22078 + }, + { + "epoch": 17.634984025559106, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 22079 + }, + { + "epoch": 17.635782747603834, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22080 + }, + { + "epoch": 17.63658146964856, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 22081 + }, + { + "epoch": 17.63738019169329, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 22082 + }, + { + "epoch": 17.63817891373802, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 22083 + }, + { + "epoch": 17.638977635782748, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 22084 + }, + { + "epoch": 17.639776357827476, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 22085 + }, + { + "epoch": 17.640575079872203, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 22086 + }, + { + "epoch": 17.641373801916934, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 22087 + }, + { + "epoch": 17.642172523961662, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 22088 + }, + { + "epoch": 17.64297124600639, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22089 + }, + { + "epoch": 17.643769968051117, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 22090 + }, + { + "epoch": 17.644568690095845, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 22091 + }, + { + "epoch": 17.645367412140576, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 22092 + }, + { + "epoch": 17.646166134185304, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 22093 + }, + { + "epoch": 17.64696485623003, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 22094 + }, + { + "epoch": 17.64776357827476, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 22095 + }, + { + "epoch": 17.64856230031949, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 22096 + }, + { + "epoch": 17.649361022364218, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 22097 + }, + { + "epoch": 17.650159744408946, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 22098 + }, + { + "epoch": 17.650958466453673, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22099 + }, + { + "epoch": 17.6517571884984, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 22100 + }, + { + "epoch": 17.652555910543132, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 22101 + }, + { + "epoch": 17.65335463258786, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 22102 + }, + { + "epoch": 17.654153354632587, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 22103 + }, + { + "epoch": 17.654952076677315, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 22104 + }, + { + "epoch": 17.655750798722046, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22105 + }, + { + "epoch": 17.656549520766774, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 22106 + }, + { + "epoch": 17.6573482428115, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 22107 + }, + { + "epoch": 17.65814696485623, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 22108 + }, + { + "epoch": 17.658945686900957, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 22109 + }, + { + "epoch": 17.659744408945688, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 22110 + }, + { + "epoch": 17.660543130990416, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22111 + }, + { + "epoch": 17.661341853035143, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 22112 + }, + { + "epoch": 17.66214057507987, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 22113 + }, + { + "epoch": 17.662939297124602, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 22114 + }, + { + "epoch": 17.66373801916933, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 22115 + }, + { + "epoch": 17.664536741214057, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 22116 + }, + { + "epoch": 17.665335463258785, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 22117 + }, + { + "epoch": 17.666134185303516, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 22118 + }, + { + "epoch": 17.666932907348244, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 22119 + }, + { + "epoch": 17.66773162939297, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 22120 + }, + { + "epoch": 17.6685303514377, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 22121 + }, + { + "epoch": 17.669329073482427, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 22122 + }, + { + "epoch": 17.670127795527158, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 22123 + }, + { + "epoch": 17.670926517571885, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 22124 + }, + { + "epoch": 17.671725239616613, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22125 + }, + { + "epoch": 17.67252396166134, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 22126 + }, + { + "epoch": 17.673322683706072, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 22127 + }, + { + "epoch": 17.6741214057508, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 22128 + }, + { + "epoch": 17.674920127795527, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 22129 + }, + { + "epoch": 17.675718849840255, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 22130 + }, + { + "epoch": 17.676517571884983, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 22131 + }, + { + "epoch": 17.677316293929714, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 22132 + }, + { + "epoch": 17.67811501597444, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 22133 + }, + { + "epoch": 17.67891373801917, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 22134 + }, + { + "epoch": 17.679712460063897, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 22135 + }, + { + "epoch": 17.680511182108628, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22136 + }, + { + "epoch": 17.681309904153355, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 22137 + }, + { + "epoch": 17.682108626198083, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 22138 + }, + { + "epoch": 17.68290734824281, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 22139 + }, + { + "epoch": 17.68370607028754, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 22140 + }, + { + "epoch": 17.68450479233227, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 22141 + }, + { + "epoch": 17.685303514376997, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 22142 + }, + { + "epoch": 17.686102236421725, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22143 + }, + { + "epoch": 17.686900958466452, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22144 + }, + { + "epoch": 17.687699680511184, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 22145 + }, + { + "epoch": 17.68849840255591, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 22146 + }, + { + "epoch": 17.68929712460064, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22147 + }, + { + "epoch": 17.690095846645367, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 22148 + }, + { + "epoch": 17.690894568690094, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 22149 + }, + { + "epoch": 17.691693290734825, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22150 + }, + { + "epoch": 17.692492012779553, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22151 + }, + { + "epoch": 17.69329073482428, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22152 + }, + { + "epoch": 17.69408945686901, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 22153 + }, + { + "epoch": 17.69488817891374, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 22154 + }, + { + "epoch": 17.695686900958467, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 22155 + }, + { + "epoch": 17.696485623003195, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22156 + }, + { + "epoch": 17.697284345047922, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 22157 + }, + { + "epoch": 17.69808306709265, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 22158 + }, + { + "epoch": 17.69888178913738, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 22159 + }, + { + "epoch": 17.69968051118211, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 22160 + }, + { + "epoch": 17.700479233226837, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 22161 + }, + { + "epoch": 17.701277955271564, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 22162 + }, + { + "epoch": 17.702076677316295, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 22163 + }, + { + "epoch": 17.702875399361023, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22164 + }, + { + "epoch": 17.70367412140575, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 22165 + }, + { + "epoch": 17.70447284345048, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 22166 + }, + { + "epoch": 17.705271565495206, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 22167 + }, + { + "epoch": 17.706070287539937, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 22168 + }, + { + "epoch": 17.706869009584665, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 22169 + }, + { + "epoch": 17.707667731629392, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 22170 + }, + { + "epoch": 17.70846645367412, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 22171 + }, + { + "epoch": 17.70926517571885, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 22172 + }, + { + "epoch": 17.71006389776358, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 22173 + }, + { + "epoch": 17.710862619808307, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 22174 + }, + { + "epoch": 17.711661341853034, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 22175 + }, + { + "epoch": 17.712460063897765, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 22176 + }, + { + "epoch": 17.713258785942493, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 22177 + }, + { + "epoch": 17.71405750798722, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 22178 + }, + { + "epoch": 17.71485623003195, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 22179 + }, + { + "epoch": 17.715654952076676, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 22180 + }, + { + "epoch": 17.716453674121407, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 22181 + }, + { + "epoch": 17.717252396166135, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 22182 + }, + { + "epoch": 17.718051118210862, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 22183 + }, + { + "epoch": 17.71884984025559, + "grad_norm": 0.095703125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 22184 + }, + { + "epoch": 17.71964856230032, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 22185 + }, + { + "epoch": 17.72044728434505, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 22186 + }, + { + "epoch": 17.721246006389777, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22187 + }, + { + "epoch": 17.722044728434504, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 22188 + }, + { + "epoch": 17.722843450479232, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 22189 + }, + { + "epoch": 17.723642172523963, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 22190 + }, + { + "epoch": 17.72444089456869, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22191 + }, + { + "epoch": 17.72523961661342, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 22192 + }, + { + "epoch": 17.726038338658146, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 22193 + }, + { + "epoch": 17.726837060702877, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 22194 + }, + { + "epoch": 17.727635782747605, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 22195 + }, + { + "epoch": 17.728434504792332, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 22196 + }, + { + "epoch": 17.72923322683706, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22197 + }, + { + "epoch": 17.730031948881788, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 22198 + }, + { + "epoch": 17.73083067092652, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22199 + }, + { + "epoch": 17.731629392971247, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 22200 + }, + { + "epoch": 17.732428115015974, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 22201 + }, + { + "epoch": 17.733226837060702, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22202 + }, + { + "epoch": 17.734025559105433, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 22203 + }, + { + "epoch": 17.73482428115016, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 22204 + }, + { + "epoch": 17.73562300319489, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 22205 + }, + { + "epoch": 17.736421725239616, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 22206 + }, + { + "epoch": 17.737220447284344, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 22207 + }, + { + "epoch": 17.738019169329075, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22208 + }, + { + "epoch": 17.738817891373802, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 22209 + }, + { + "epoch": 17.73961661341853, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22210 + }, + { + "epoch": 17.740415335463258, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 22211 + }, + { + "epoch": 17.74121405750799, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 22212 + }, + { + "epoch": 17.742012779552716, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 22213 + }, + { + "epoch": 17.742811501597444, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 22214 + }, + { + "epoch": 17.74361022364217, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 22215 + }, + { + "epoch": 17.7444089456869, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 22216 + }, + { + "epoch": 17.74520766773163, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 22217 + }, + { + "epoch": 17.74600638977636, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 22218 + }, + { + "epoch": 17.746805111821086, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 22219 + }, + { + "epoch": 17.747603833865814, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 22220 + }, + { + "epoch": 17.748402555910545, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 22221 + }, + { + "epoch": 17.749201277955272, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 22222 + }, + { + "epoch": 17.75, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 22223 + }, + { + "epoch": 17.750798722044728, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 22224 + }, + { + "epoch": 17.751597444089455, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 22225 + }, + { + "epoch": 17.752396166134186, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 22226 + }, + { + "epoch": 17.753194888178914, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 22227 + }, + { + "epoch": 17.75399361022364, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 22228 + }, + { + "epoch": 17.75479233226837, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 22229 + }, + { + "epoch": 17.7555910543131, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 22230 + }, + { + "epoch": 17.75638977635783, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 22231 + }, + { + "epoch": 17.757188498402556, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 22232 + }, + { + "epoch": 17.757987220447284, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 22233 + }, + { + "epoch": 17.75878594249201, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 22234 + }, + { + "epoch": 17.759584664536742, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 22235 + }, + { + "epoch": 17.76038338658147, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 22236 + }, + { + "epoch": 17.761182108626198, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 22237 + }, + { + "epoch": 17.761980830670925, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 22238 + }, + { + "epoch": 17.762779552715656, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 22239 + }, + { + "epoch": 17.763578274760384, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 22240 + }, + { + "epoch": 17.76437699680511, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 22241 + }, + { + "epoch": 17.76517571884984, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22242 + }, + { + "epoch": 17.765974440894567, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 22243 + }, + { + "epoch": 17.766773162939298, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 22244 + }, + { + "epoch": 17.767571884984026, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 22245 + }, + { + "epoch": 17.768370607028753, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 22246 + }, + { + "epoch": 17.76916932907348, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 22247 + }, + { + "epoch": 17.769968051118212, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 22248 + }, + { + "epoch": 17.77076677316294, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22249 + }, + { + "epoch": 17.771565495207668, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22250 + }, + { + "epoch": 17.772364217252395, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 22251 + }, + { + "epoch": 17.773162939297123, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 22252 + }, + { + "epoch": 17.773961661341854, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 22253 + }, + { + "epoch": 17.77476038338658, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 22254 + }, + { + "epoch": 17.77555910543131, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 22255 + }, + { + "epoch": 17.776357827476037, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 22256 + }, + { + "epoch": 17.777156549520768, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 22257 + }, + { + "epoch": 17.777955271565496, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22258 + }, + { + "epoch": 17.778753993610223, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 22259 + }, + { + "epoch": 17.77955271565495, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 22260 + }, + { + "epoch": 17.78035143769968, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 22261 + }, + { + "epoch": 17.78115015974441, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 22262 + }, + { + "epoch": 17.781948881789138, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 22263 + }, + { + "epoch": 17.782747603833865, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 22264 + }, + { + "epoch": 17.783546325878593, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 22265 + }, + { + "epoch": 17.784345047923324, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 22266 + }, + { + "epoch": 17.78514376996805, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 22267 + }, + { + "epoch": 17.78594249201278, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 22268 + }, + { + "epoch": 17.786741214057507, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 22269 + }, + { + "epoch": 17.787539936102235, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 22270 + }, + { + "epoch": 17.788338658146966, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 22271 + }, + { + "epoch": 17.789137380191693, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 22272 + }, + { + "epoch": 17.78993610223642, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 22273 + }, + { + "epoch": 17.79073482428115, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 22274 + }, + { + "epoch": 17.79153354632588, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 22275 + }, + { + "epoch": 17.792332268370608, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22276 + }, + { + "epoch": 17.793130990415335, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 22277 + }, + { + "epoch": 17.793929712460063, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22278 + }, + { + "epoch": 17.794728434504794, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 22279 + }, + { + "epoch": 17.79552715654952, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 22280 + }, + { + "epoch": 17.79632587859425, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 22281 + }, + { + "epoch": 17.797124600638977, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 22282 + }, + { + "epoch": 17.797923322683705, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22283 + }, + { + "epoch": 17.798722044728436, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22284 + }, + { + "epoch": 17.799520766773163, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 22285 + }, + { + "epoch": 17.80031948881789, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 22286 + }, + { + "epoch": 17.80111821086262, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22287 + }, + { + "epoch": 17.80191693290735, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22288 + }, + { + "epoch": 17.802715654952078, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 22289 + }, + { + "epoch": 17.803514376996805, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 22290 + }, + { + "epoch": 17.804313099041533, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 22291 + }, + { + "epoch": 17.80511182108626, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 22292 + }, + { + "epoch": 17.80591054313099, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 22293 + }, + { + "epoch": 17.80670926517572, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 22294 + }, + { + "epoch": 17.807507987220447, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 22295 + }, + { + "epoch": 17.808306709265175, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 22296 + }, + { + "epoch": 17.809105431309906, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 22297 + }, + { + "epoch": 17.809904153354633, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22298 + }, + { + "epoch": 17.81070287539936, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 22299 + }, + { + "epoch": 17.81150159744409, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 22300 + }, + { + "epoch": 17.812300319488816, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 22301 + }, + { + "epoch": 17.813099041533548, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 22302 + }, + { + "epoch": 17.813897763578275, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 22303 + }, + { + "epoch": 17.814696485623003, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 22304 + }, + { + "epoch": 17.81549520766773, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 22305 + }, + { + "epoch": 17.81629392971246, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 22306 + }, + { + "epoch": 17.81709265175719, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 22307 + }, + { + "epoch": 17.817891373801917, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22308 + }, + { + "epoch": 17.818690095846645, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 22309 + }, + { + "epoch": 17.819488817891372, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 22310 + }, + { + "epoch": 17.820287539936103, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 22311 + }, + { + "epoch": 17.82108626198083, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 22312 + }, + { + "epoch": 17.82188498402556, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 22313 + }, + { + "epoch": 17.822683706070286, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 22314 + }, + { + "epoch": 17.823482428115017, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 22315 + }, + { + "epoch": 17.824281150159745, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 22316 + }, + { + "epoch": 17.825079872204473, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 22317 + }, + { + "epoch": 17.8258785942492, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 22318 + }, + { + "epoch": 17.826677316293928, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 22319 + }, + { + "epoch": 17.82747603833866, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 22320 + }, + { + "epoch": 17.828274760383387, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 22321 + }, + { + "epoch": 17.829073482428115, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 22322 + }, + { + "epoch": 17.829872204472842, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 22323 + }, + { + "epoch": 17.830670926517573, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22324 + }, + { + "epoch": 17.8314696485623, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 22325 + }, + { + "epoch": 17.83226837060703, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 22326 + }, + { + "epoch": 17.833067092651756, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 22327 + }, + { + "epoch": 17.833865814696484, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22328 + }, + { + "epoch": 17.834664536741215, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 22329 + }, + { + "epoch": 17.835463258785943, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 22330 + }, + { + "epoch": 17.83626198083067, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 22331 + }, + { + "epoch": 17.837060702875398, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 22332 + }, + { + "epoch": 17.83785942492013, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 22333 + }, + { + "epoch": 17.838658146964857, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 22334 + }, + { + "epoch": 17.839456869009584, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 22335 + }, + { + "epoch": 17.840255591054312, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22336 + }, + { + "epoch": 17.841054313099043, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 22337 + }, + { + "epoch": 17.84185303514377, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22338 + }, + { + "epoch": 17.8426517571885, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22339 + }, + { + "epoch": 17.843450479233226, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 22340 + }, + { + "epoch": 17.844249201277954, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 22341 + }, + { + "epoch": 17.845047923322685, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 22342 + }, + { + "epoch": 17.845846645367413, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 22343 + }, + { + "epoch": 17.84664536741214, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 22344 + }, + { + "epoch": 17.847444089456868, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 22345 + }, + { + "epoch": 17.8482428115016, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22346 + }, + { + "epoch": 17.849041533546327, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 22347 + }, + { + "epoch": 17.849840255591054, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 22348 + }, + { + "epoch": 17.850638977635782, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 22349 + }, + { + "epoch": 17.85143769968051, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 22350 + }, + { + "epoch": 17.85223642172524, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 22351 + }, + { + "epoch": 17.85303514376997, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 22352 + }, + { + "epoch": 17.853833865814696, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 22353 + }, + { + "epoch": 17.854632587859424, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 22354 + }, + { + "epoch": 17.855431309904155, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 22355 + }, + { + "epoch": 17.856230031948883, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 22356 + }, + { + "epoch": 17.85702875399361, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 22357 + }, + { + "epoch": 17.857827476038338, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 22358 + }, + { + "epoch": 17.858626198083066, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 22359 + }, + { + "epoch": 17.859424920127797, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 22360 + }, + { + "epoch": 17.860223642172524, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 22361 + }, + { + "epoch": 17.861022364217252, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 22362 + }, + { + "epoch": 17.86182108626198, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 22363 + }, + { + "epoch": 17.86261980830671, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 22364 + }, + { + "epoch": 17.86341853035144, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 22365 + }, + { + "epoch": 17.864217252396166, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 22366 + }, + { + "epoch": 17.865015974440894, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 22367 + }, + { + "epoch": 17.86581469648562, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 22368 + }, + { + "epoch": 17.866613418530353, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 22369 + }, + { + "epoch": 17.86741214057508, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 22370 + }, + { + "epoch": 17.868210862619808, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 22371 + }, + { + "epoch": 17.869009584664536, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22372 + }, + { + "epoch": 17.869808306709267, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 22373 + }, + { + "epoch": 17.870607028753994, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 22374 + }, + { + "epoch": 17.871405750798722, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 22375 + }, + { + "epoch": 17.87220447284345, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 22376 + }, + { + "epoch": 17.873003194888177, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22377 + }, + { + "epoch": 17.87380191693291, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22378 + }, + { + "epoch": 17.874600638977636, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 22379 + }, + { + "epoch": 17.875399361022364, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 22380 + }, + { + "epoch": 17.87619808306709, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22381 + }, + { + "epoch": 17.876996805111823, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 22382 + }, + { + "epoch": 17.87779552715655, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 22383 + }, + { + "epoch": 17.878594249201278, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 22384 + }, + { + "epoch": 17.879392971246006, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 22385 + }, + { + "epoch": 17.880191693290733, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 22386 + }, + { + "epoch": 17.880990415335464, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 22387 + }, + { + "epoch": 17.881789137380192, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 22388 + }, + { + "epoch": 17.88258785942492, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 22389 + }, + { + "epoch": 17.883386581469647, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 22390 + }, + { + "epoch": 17.88418530351438, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 22391 + }, + { + "epoch": 17.884984025559106, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 22392 + }, + { + "epoch": 17.885782747603834, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 22393 + }, + { + "epoch": 17.88658146964856, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 22394 + }, + { + "epoch": 17.88738019169329, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22395 + }, + { + "epoch": 17.88817891373802, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 22396 + }, + { + "epoch": 17.888977635782748, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 22397 + }, + { + "epoch": 17.889776357827476, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 22398 + }, + { + "epoch": 17.890575079872203, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 22399 + }, + { + "epoch": 17.891373801916934, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 22400 + }, + { + "epoch": 17.892172523961662, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 22401 + }, + { + "epoch": 17.89297124600639, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 22402 + }, + { + "epoch": 17.893769968051117, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 22403 + }, + { + "epoch": 17.894568690095845, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 22404 + }, + { + "epoch": 17.895367412140576, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22405 + }, + { + "epoch": 17.896166134185304, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 22406 + }, + { + "epoch": 17.89696485623003, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 22407 + }, + { + "epoch": 17.89776357827476, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 22408 + }, + { + "epoch": 17.89856230031949, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 22409 + }, + { + "epoch": 17.899361022364218, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 22410 + }, + { + "epoch": 17.900159744408946, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 22411 + }, + { + "epoch": 17.900958466453673, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 22412 + }, + { + "epoch": 17.9017571884984, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 22413 + }, + { + "epoch": 17.902555910543132, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 22414 + }, + { + "epoch": 17.90335463258786, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22415 + }, + { + "epoch": 17.904153354632587, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 22416 + }, + { + "epoch": 17.904952076677315, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 22417 + }, + { + "epoch": 17.905750798722046, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 22418 + }, + { + "epoch": 17.906549520766774, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 22419 + }, + { + "epoch": 17.9073482428115, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 22420 + }, + { + "epoch": 17.90814696485623, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 22421 + }, + { + "epoch": 17.908945686900957, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 22422 + }, + { + "epoch": 17.909744408945688, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 22423 + }, + { + "epoch": 17.910543130990416, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 22424 + }, + { + "epoch": 17.911341853035143, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 22425 + }, + { + "epoch": 17.91214057507987, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 22426 + }, + { + "epoch": 17.912939297124602, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 22427 + }, + { + "epoch": 17.91373801916933, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 22428 + }, + { + "epoch": 17.914536741214057, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 22429 + }, + { + "epoch": 17.915335463258785, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 22430 + }, + { + "epoch": 17.916134185303516, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 22431 + }, + { + "epoch": 17.916932907348244, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 22432 + }, + { + "epoch": 17.91773162939297, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 22433 + }, + { + "epoch": 17.9185303514377, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22434 + }, + { + "epoch": 17.919329073482427, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 22435 + }, + { + "epoch": 17.920127795527158, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 22436 + }, + { + "epoch": 17.920926517571885, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 22437 + }, + { + "epoch": 17.921725239616613, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 22438 + }, + { + "epoch": 17.92252396166134, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 22439 + }, + { + "epoch": 17.923322683706072, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 22440 + }, + { + "epoch": 17.9241214057508, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 22441 + }, + { + "epoch": 17.924920127795527, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 22442 + }, + { + "epoch": 17.925718849840255, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 22443 + }, + { + "epoch": 17.926517571884983, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 22444 + }, + { + "epoch": 17.927316293929714, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 22445 + }, + { + "epoch": 17.92811501597444, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 22446 + }, + { + "epoch": 17.92891373801917, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 22447 + }, + { + "epoch": 17.929712460063897, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22448 + }, + { + "epoch": 17.930511182108628, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 22449 + }, + { + "epoch": 17.931309904153355, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 22450 + }, + { + "epoch": 17.932108626198083, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 22451 + }, + { + "epoch": 17.93290734824281, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 22452 + }, + { + "epoch": 17.93370607028754, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 22453 + }, + { + "epoch": 17.93450479233227, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22454 + }, + { + "epoch": 17.935303514376997, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 22455 + }, + { + "epoch": 17.936102236421725, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 22456 + }, + { + "epoch": 17.936900958466452, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22457 + }, + { + "epoch": 17.937699680511184, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 22458 + }, + { + "epoch": 17.93849840255591, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 22459 + }, + { + "epoch": 17.93929712460064, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 22460 + }, + { + "epoch": 17.940095846645367, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 22461 + }, + { + "epoch": 17.940894568690094, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 22462 + }, + { + "epoch": 17.941693290734825, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 22463 + }, + { + "epoch": 17.942492012779553, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 22464 + }, + { + "epoch": 17.94329073482428, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 22465 + }, + { + "epoch": 17.94408945686901, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 22466 + }, + { + "epoch": 17.94488817891374, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 22467 + }, + { + "epoch": 17.945686900958467, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 22468 + }, + { + "epoch": 17.946485623003195, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 22469 + }, + { + "epoch": 17.947284345047922, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 22470 + }, + { + "epoch": 17.94808306709265, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 22471 + }, + { + "epoch": 17.94888178913738, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 22472 + }, + { + "epoch": 17.94968051118211, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22473 + }, + { + "epoch": 17.950479233226837, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 22474 + }, + { + "epoch": 17.951277955271564, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 22475 + }, + { + "epoch": 17.952076677316295, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22476 + }, + { + "epoch": 17.952875399361023, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 22477 + }, + { + "epoch": 17.95367412140575, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 22478 + }, + { + "epoch": 17.95447284345048, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 22479 + }, + { + "epoch": 17.955271565495206, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 22480 + }, + { + "epoch": 17.956070287539937, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 22481 + }, + { + "epoch": 17.956869009584665, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 22482 + }, + { + "epoch": 17.957667731629392, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 22483 + }, + { + "epoch": 17.95846645367412, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 22484 + }, + { + "epoch": 17.95926517571885, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 22485 + }, + { + "epoch": 17.96006389776358, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 22486 + }, + { + "epoch": 17.960862619808307, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 22487 + }, + { + "epoch": 17.961661341853034, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 22488 + }, + { + "epoch": 17.962460063897765, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 22489 + }, + { + "epoch": 17.963258785942493, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 22490 + }, + { + "epoch": 17.96405750798722, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 22491 + }, + { + "epoch": 17.96485623003195, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22492 + }, + { + "epoch": 17.965654952076676, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 22493 + }, + { + "epoch": 17.966453674121407, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 22494 + }, + { + "epoch": 17.967252396166135, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 22495 + }, + { + "epoch": 17.968051118210862, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 22496 + }, + { + "epoch": 17.96884984025559, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22497 + }, + { + "epoch": 17.96964856230032, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 22498 + }, + { + "epoch": 17.97044728434505, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 22499 + }, + { + "epoch": 17.971246006389777, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 22500 + }, + { + "epoch": 17.972044728434504, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 22501 + }, + { + "epoch": 17.972843450479232, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 22502 + }, + { + "epoch": 17.973642172523963, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 22503 + }, + { + "epoch": 17.97444089456869, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 22504 + }, + { + "epoch": 17.97523961661342, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 22505 + }, + { + "epoch": 17.976038338658146, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 22506 + }, + { + "epoch": 17.976837060702877, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 22507 + }, + { + "epoch": 17.977635782747605, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22508 + }, + { + "epoch": 17.978434504792332, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 22509 + }, + { + "epoch": 17.97923322683706, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 22510 + }, + { + "epoch": 17.980031948881788, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 22511 + }, + { + "epoch": 17.98083067092652, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 22512 + }, + { + "epoch": 17.981629392971247, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 22513 + }, + { + "epoch": 17.982428115015974, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 22514 + }, + { + "epoch": 17.983226837060702, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 22515 + }, + { + "epoch": 17.984025559105433, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 22516 + }, + { + "epoch": 17.98482428115016, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 22517 + }, + { + "epoch": 17.98562300319489, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 22518 + }, + { + "epoch": 17.986421725239616, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 22519 + }, + { + "epoch": 17.987220447284344, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22520 + }, + { + "epoch": 17.988019169329075, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 22521 + }, + { + "epoch": 17.988817891373802, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 22522 + }, + { + "epoch": 17.98961661341853, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 22523 + }, + { + "epoch": 17.990415335463258, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 22524 + }, + { + "epoch": 17.99121405750799, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 22525 + }, + { + "epoch": 17.992012779552716, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 22526 + }, + { + "epoch": 17.992811501597444, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 22527 + }, + { + "epoch": 17.99361022364217, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 22528 + }, + { + "epoch": 17.9944089456869, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 22529 + }, + { + "epoch": 17.99520766773163, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 22530 + }, + { + "epoch": 17.99600638977636, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 22531 + }, + { + "epoch": 17.996805111821086, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 22532 + }, + { + "epoch": 17.997603833865814, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 22533 + }, + { + "epoch": 17.998402555910545, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 22534 + }, + { + "epoch": 17.999201277955272, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 22535 + }, + { + "epoch": 18.0, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 22536 + }, + { + "epoch": 18.000798722044728, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 22537 + }, + { + "epoch": 18.001597444089455, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 22538 + }, + { + "epoch": 18.002396166134186, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 22539 + }, + { + "epoch": 18.003194888178914, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 22540 + }, + { + "epoch": 18.00399361022364, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 22541 + }, + { + "epoch": 18.00479233226837, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 22542 + }, + { + "epoch": 18.0055910543131, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 22543 + }, + { + "epoch": 18.00638977635783, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 22544 + }, + { + "epoch": 18.007188498402556, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 22545 + }, + { + "epoch": 18.007987220447284, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 22546 + }, + { + "epoch": 18.00878594249201, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 22547 + }, + { + "epoch": 18.009584664536742, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 22548 + }, + { + "epoch": 18.01038338658147, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 22549 + }, + { + "epoch": 18.011182108626198, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 22550 + }, + { + "epoch": 18.011980830670925, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 22551 + }, + { + "epoch": 18.012779552715656, + "grad_norm": 0.328125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 22552 + }, + { + "epoch": 18.013578274760384, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 22553 + }, + { + "epoch": 18.01437699680511, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 22554 + }, + { + "epoch": 18.01517571884984, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 22555 + }, + { + "epoch": 18.015974440894567, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22556 + }, + { + "epoch": 18.016773162939298, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 22557 + }, + { + "epoch": 18.017571884984026, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 22558 + }, + { + "epoch": 18.018370607028753, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 22559 + }, + { + "epoch": 18.01916932907348, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 22560 + }, + { + "epoch": 18.019968051118212, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 22561 + }, + { + "epoch": 18.02076677316294, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 22562 + }, + { + "epoch": 18.021565495207668, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 22563 + }, + { + "epoch": 18.022364217252395, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 22564 + }, + { + "epoch": 18.023162939297123, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 22565 + }, + { + "epoch": 18.023961661341854, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 22566 + }, + { + "epoch": 18.02476038338658, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 22567 + }, + { + "epoch": 18.02555910543131, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 22568 + }, + { + "epoch": 18.026357827476037, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 22569 + }, + { + "epoch": 18.027156549520768, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 22570 + }, + { + "epoch": 18.027955271565496, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 22571 + }, + { + "epoch": 18.028753993610223, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 22572 + }, + { + "epoch": 18.02955271565495, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 22573 + }, + { + "epoch": 18.03035143769968, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 22574 + }, + { + "epoch": 18.03115015974441, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 22575 + }, + { + "epoch": 18.031948881789138, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 22576 + }, + { + "epoch": 18.032747603833865, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 22577 + }, + { + "epoch": 18.033546325878593, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22578 + }, + { + "epoch": 18.034345047923324, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 22579 + }, + { + "epoch": 18.03514376996805, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 22580 + }, + { + "epoch": 18.03594249201278, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 22581 + }, + { + "epoch": 18.036741214057507, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22582 + }, + { + "epoch": 18.037539936102238, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 22583 + }, + { + "epoch": 18.038338658146966, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 22584 + }, + { + "epoch": 18.039137380191693, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 22585 + }, + { + "epoch": 18.03993610223642, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 22586 + }, + { + "epoch": 18.04073482428115, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 22587 + }, + { + "epoch": 18.04153354632588, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 22588 + }, + { + "epoch": 18.042332268370608, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 22589 + }, + { + "epoch": 18.043130990415335, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 22590 + }, + { + "epoch": 18.043929712460063, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 22591 + }, + { + "epoch": 18.044728434504794, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 22592 + }, + { + "epoch": 18.04552715654952, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22593 + }, + { + "epoch": 18.04632587859425, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 22594 + }, + { + "epoch": 18.047124600638977, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 22595 + }, + { + "epoch": 18.047923322683705, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 22596 + }, + { + "epoch": 18.048722044728436, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 22597 + }, + { + "epoch": 18.049520766773163, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22598 + }, + { + "epoch": 18.05031948881789, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 22599 + }, + { + "epoch": 18.05111821086262, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22600 + }, + { + "epoch": 18.05191693290735, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 22601 + }, + { + "epoch": 18.052715654952078, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 22602 + }, + { + "epoch": 18.053514376996805, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 22603 + }, + { + "epoch": 18.054313099041533, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 22604 + }, + { + "epoch": 18.05511182108626, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22605 + }, + { + "epoch": 18.05591054313099, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 22606 + }, + { + "epoch": 18.05670926517572, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 22607 + }, + { + "epoch": 18.057507987220447, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22608 + }, + { + "epoch": 18.058306709265175, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 22609 + }, + { + "epoch": 18.059105431309906, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 22610 + }, + { + "epoch": 18.059904153354633, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 22611 + }, + { + "epoch": 18.06070287539936, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 22612 + }, + { + "epoch": 18.06150159744409, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 22613 + }, + { + "epoch": 18.062300319488816, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 22614 + }, + { + "epoch": 18.063099041533548, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22615 + }, + { + "epoch": 18.063897763578275, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 22616 + }, + { + "epoch": 18.064696485623003, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22617 + }, + { + "epoch": 18.06549520766773, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 22618 + }, + { + "epoch": 18.06629392971246, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 22619 + }, + { + "epoch": 18.06709265175719, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 22620 + }, + { + "epoch": 18.067891373801917, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 22621 + }, + { + "epoch": 18.068690095846645, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 22622 + }, + { + "epoch": 18.069488817891372, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 22623 + }, + { + "epoch": 18.070287539936103, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 22624 + }, + { + "epoch": 18.07108626198083, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 22625 + }, + { + "epoch": 18.07188498402556, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 22626 + }, + { + "epoch": 18.072683706070286, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 22627 + }, + { + "epoch": 18.073482428115017, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 22628 + }, + { + "epoch": 18.074281150159745, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 22629 + }, + { + "epoch": 18.075079872204473, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 22630 + }, + { + "epoch": 18.0758785942492, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 22631 + }, + { + "epoch": 18.076677316293928, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 22632 + }, + { + "epoch": 18.07747603833866, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 22633 + }, + { + "epoch": 18.078274760383387, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 22634 + }, + { + "epoch": 18.079073482428115, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 22635 + }, + { + "epoch": 18.079872204472842, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 22636 + }, + { + "epoch": 18.080670926517573, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 22637 + }, + { + "epoch": 18.0814696485623, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 22638 + }, + { + "epoch": 18.08226837060703, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 22639 + }, + { + "epoch": 18.083067092651756, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 22640 + }, + { + "epoch": 18.083865814696484, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 22641 + }, + { + "epoch": 18.084664536741215, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 22642 + }, + { + "epoch": 18.085463258785943, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 22643 + }, + { + "epoch": 18.08626198083067, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 22644 + }, + { + "epoch": 18.087060702875398, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22645 + }, + { + "epoch": 18.08785942492013, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 22646 + }, + { + "epoch": 18.088658146964857, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 22647 + }, + { + "epoch": 18.089456869009584, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 22648 + }, + { + "epoch": 18.090255591054312, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 22649 + }, + { + "epoch": 18.09105431309904, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22650 + }, + { + "epoch": 18.09185303514377, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 22651 + }, + { + "epoch": 18.0926517571885, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 22652 + }, + { + "epoch": 18.093450479233226, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 22653 + }, + { + "epoch": 18.094249201277954, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 22654 + }, + { + "epoch": 18.095047923322685, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1057, + "step": 22655 + }, + { + "epoch": 18.095846645367413, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 22656 + }, + { + "epoch": 18.09664536741214, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 22657 + }, + { + "epoch": 18.097444089456868, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 22658 + }, + { + "epoch": 18.0982428115016, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 22659 + }, + { + "epoch": 18.099041533546327, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 22660 + }, + { + "epoch": 18.099840255591054, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 22661 + }, + { + "epoch": 18.100638977635782, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 22662 + }, + { + "epoch": 18.10143769968051, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 22663 + }, + { + "epoch": 18.10223642172524, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 22664 + }, + { + "epoch": 18.10303514376997, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 22665 + }, + { + "epoch": 18.103833865814696, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 22666 + }, + { + "epoch": 18.104632587859424, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 22667 + }, + { + "epoch": 18.105431309904155, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 22668 + }, + { + "epoch": 18.106230031948883, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 22669 + }, + { + "epoch": 18.10702875399361, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 22670 + }, + { + "epoch": 18.107827476038338, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 22671 + }, + { + "epoch": 18.108626198083066, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 22672 + }, + { + "epoch": 18.109424920127797, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 22673 + }, + { + "epoch": 18.110223642172524, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 22674 + }, + { + "epoch": 18.111022364217252, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22675 + }, + { + "epoch": 18.11182108626198, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 22676 + }, + { + "epoch": 18.11261980830671, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 22677 + }, + { + "epoch": 18.11341853035144, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 22678 + }, + { + "epoch": 18.114217252396166, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 22679 + }, + { + "epoch": 18.115015974440894, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 22680 + }, + { + "epoch": 18.11581469648562, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 22681 + }, + { + "epoch": 18.116613418530353, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 22682 + }, + { + "epoch": 18.11741214057508, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 22683 + }, + { + "epoch": 18.118210862619808, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 22684 + }, + { + "epoch": 18.119009584664536, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 22685 + }, + { + "epoch": 18.119808306709267, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22686 + }, + { + "epoch": 18.120607028753994, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 22687 + }, + { + "epoch": 18.121405750798722, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 22688 + }, + { + "epoch": 18.12220447284345, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 22689 + }, + { + "epoch": 18.123003194888177, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 22690 + }, + { + "epoch": 18.12380191693291, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22691 + }, + { + "epoch": 18.124600638977636, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22692 + }, + { + "epoch": 18.125399361022364, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 22693 + }, + { + "epoch": 18.12619808306709, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 22694 + }, + { + "epoch": 18.126996805111823, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 22695 + }, + { + "epoch": 18.12779552715655, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22696 + }, + { + "epoch": 18.128594249201278, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 22697 + }, + { + "epoch": 18.129392971246006, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 22698 + }, + { + "epoch": 18.130191693290733, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 22699 + }, + { + "epoch": 18.130990415335464, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22700 + }, + { + "epoch": 18.131789137380192, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 22701 + }, + { + "epoch": 18.13258785942492, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22702 + }, + { + "epoch": 18.133386581469647, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 22703 + }, + { + "epoch": 18.13418530351438, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 22704 + }, + { + "epoch": 18.134984025559106, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 22705 + }, + { + "epoch": 18.135782747603834, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 22706 + }, + { + "epoch": 18.13658146964856, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 22707 + }, + { + "epoch": 18.13738019169329, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 22708 + }, + { + "epoch": 18.13817891373802, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 22709 + }, + { + "epoch": 18.138977635782748, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 22710 + }, + { + "epoch": 18.139776357827476, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 22711 + }, + { + "epoch": 18.140575079872203, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 22712 + }, + { + "epoch": 18.141373801916934, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 22713 + }, + { + "epoch": 18.142172523961662, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 22714 + }, + { + "epoch": 18.14297124600639, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 22715 + }, + { + "epoch": 18.143769968051117, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 22716 + }, + { + "epoch": 18.144568690095845, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 22717 + }, + { + "epoch": 18.145367412140576, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 22718 + }, + { + "epoch": 18.146166134185304, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 22719 + }, + { + "epoch": 18.14696485623003, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22720 + }, + { + "epoch": 18.14776357827476, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 22721 + }, + { + "epoch": 18.14856230031949, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 22722 + }, + { + "epoch": 18.149361022364218, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22723 + }, + { + "epoch": 18.150159744408946, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 22724 + }, + { + "epoch": 18.150958466453673, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 22725 + }, + { + "epoch": 18.1517571884984, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22726 + }, + { + "epoch": 18.152555910543132, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 22727 + }, + { + "epoch": 18.15335463258786, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 22728 + }, + { + "epoch": 18.154153354632587, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 22729 + }, + { + "epoch": 18.154952076677315, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 22730 + }, + { + "epoch": 18.155750798722046, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 22731 + }, + { + "epoch": 18.156549520766774, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22732 + }, + { + "epoch": 18.1573482428115, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 22733 + }, + { + "epoch": 18.15814696485623, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 22734 + }, + { + "epoch": 18.15894568690096, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 22735 + }, + { + "epoch": 18.159744408945688, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22736 + }, + { + "epoch": 18.160543130990416, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 22737 + }, + { + "epoch": 18.161341853035143, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22738 + }, + { + "epoch": 18.16214057507987, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 22739 + }, + { + "epoch": 18.162939297124602, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 22740 + }, + { + "epoch": 18.16373801916933, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 22741 + }, + { + "epoch": 18.164536741214057, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 22742 + }, + { + "epoch": 18.165335463258785, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 22743 + }, + { + "epoch": 18.166134185303516, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 22744 + }, + { + "epoch": 18.166932907348244, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 22745 + }, + { + "epoch": 18.16773162939297, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 22746 + }, + { + "epoch": 18.1685303514377, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 22747 + }, + { + "epoch": 18.169329073482427, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 22748 + }, + { + "epoch": 18.170127795527158, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 22749 + }, + { + "epoch": 18.170926517571885, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22750 + }, + { + "epoch": 18.171725239616613, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 22751 + }, + { + "epoch": 18.17252396166134, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 22752 + }, + { + "epoch": 18.173322683706072, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 22753 + }, + { + "epoch": 18.1741214057508, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 22754 + }, + { + "epoch": 18.174920127795527, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 22755 + }, + { + "epoch": 18.175718849840255, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 22756 + }, + { + "epoch": 18.176517571884983, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 22757 + }, + { + "epoch": 18.177316293929714, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 22758 + }, + { + "epoch": 18.17811501597444, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 22759 + }, + { + "epoch": 18.17891373801917, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 22760 + }, + { + "epoch": 18.179712460063897, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22761 + }, + { + "epoch": 18.180511182108628, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 22762 + }, + { + "epoch": 18.181309904153355, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 22763 + }, + { + "epoch": 18.182108626198083, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22764 + }, + { + "epoch": 18.18290734824281, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 22765 + }, + { + "epoch": 18.18370607028754, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 22766 + }, + { + "epoch": 18.18450479233227, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 22767 + }, + { + "epoch": 18.185303514376997, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 22768 + }, + { + "epoch": 18.186102236421725, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 22769 + }, + { + "epoch": 18.186900958466452, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 22770 + }, + { + "epoch": 18.187699680511184, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 22771 + }, + { + "epoch": 18.18849840255591, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 22772 + }, + { + "epoch": 18.18929712460064, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 22773 + }, + { + "epoch": 18.190095846645367, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22774 + }, + { + "epoch": 18.190894568690094, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 22775 + }, + { + "epoch": 18.191693290734825, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 22776 + }, + { + "epoch": 18.192492012779553, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 22777 + }, + { + "epoch": 18.19329073482428, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 22778 + }, + { + "epoch": 18.19408945686901, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22779 + }, + { + "epoch": 18.19488817891374, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 22780 + }, + { + "epoch": 18.195686900958467, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 22781 + }, + { + "epoch": 18.196485623003195, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 22782 + }, + { + "epoch": 18.197284345047922, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 22783 + }, + { + "epoch": 18.19808306709265, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 22784 + }, + { + "epoch": 18.19888178913738, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 22785 + }, + { + "epoch": 18.19968051118211, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 22786 + }, + { + "epoch": 18.200479233226837, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 22787 + }, + { + "epoch": 18.201277955271564, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 22788 + }, + { + "epoch": 18.202076677316295, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 22789 + }, + { + "epoch": 18.202875399361023, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 22790 + }, + { + "epoch": 18.20367412140575, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22791 + }, + { + "epoch": 18.20447284345048, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 22792 + }, + { + "epoch": 18.205271565495206, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 22793 + }, + { + "epoch": 18.206070287539937, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 22794 + }, + { + "epoch": 18.206869009584665, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 22795 + }, + { + "epoch": 18.207667731629392, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 22796 + }, + { + "epoch": 18.20846645367412, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 22797 + }, + { + "epoch": 18.20926517571885, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 22798 + }, + { + "epoch": 18.21006389776358, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 22799 + }, + { + "epoch": 18.210862619808307, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 22800 + }, + { + "epoch": 18.211661341853034, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 22801 + }, + { + "epoch": 18.212460063897762, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 22802 + }, + { + "epoch": 18.213258785942493, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 22803 + }, + { + "epoch": 18.21405750798722, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 22804 + }, + { + "epoch": 18.21485623003195, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 22805 + }, + { + "epoch": 18.215654952076676, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 22806 + }, + { + "epoch": 18.216453674121407, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 22807 + }, + { + "epoch": 18.217252396166135, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 22808 + }, + { + "epoch": 18.218051118210862, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 22809 + }, + { + "epoch": 18.21884984025559, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 22810 + }, + { + "epoch": 18.21964856230032, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 22811 + }, + { + "epoch": 18.22044728434505, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 22812 + }, + { + "epoch": 18.221246006389777, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22813 + }, + { + "epoch": 18.222044728434504, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 22814 + }, + { + "epoch": 18.222843450479232, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 22815 + }, + { + "epoch": 18.223642172523963, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 22816 + }, + { + "epoch": 18.22444089456869, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 22817 + }, + { + "epoch": 18.22523961661342, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 22818 + }, + { + "epoch": 18.226038338658146, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22819 + }, + { + "epoch": 18.226837060702877, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 22820 + }, + { + "epoch": 18.227635782747605, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22821 + }, + { + "epoch": 18.228434504792332, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 22822 + }, + { + "epoch": 18.22923322683706, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 22823 + }, + { + "epoch": 18.230031948881788, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 22824 + }, + { + "epoch": 18.23083067092652, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 22825 + }, + { + "epoch": 18.231629392971247, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 22826 + }, + { + "epoch": 18.232428115015974, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 22827 + }, + { + "epoch": 18.233226837060702, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 22828 + }, + { + "epoch": 18.234025559105433, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 22829 + }, + { + "epoch": 18.23482428115016, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 22830 + }, + { + "epoch": 18.23562300319489, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 22831 + }, + { + "epoch": 18.236421725239616, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 22832 + }, + { + "epoch": 18.237220447284344, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 22833 + }, + { + "epoch": 18.238019169329075, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 22834 + }, + { + "epoch": 18.238817891373802, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 22835 + }, + { + "epoch": 18.23961661341853, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 22836 + }, + { + "epoch": 18.240415335463258, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 22837 + }, + { + "epoch": 18.24121405750799, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 22838 + }, + { + "epoch": 18.242012779552716, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 22839 + }, + { + "epoch": 18.242811501597444, + "grad_norm": 0.0966796875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22840 + }, + { + "epoch": 18.24361022364217, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22841 + }, + { + "epoch": 18.2444089456869, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 22842 + }, + { + "epoch": 18.24520766773163, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22843 + }, + { + "epoch": 18.24600638977636, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 22844 + }, + { + "epoch": 18.246805111821086, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 22845 + }, + { + "epoch": 18.247603833865814, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 22846 + }, + { + "epoch": 18.248402555910545, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 22847 + }, + { + "epoch": 18.249201277955272, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 22848 + }, + { + "epoch": 18.25, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 22849 + }, + { + "epoch": 18.250798722044728, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 22850 + }, + { + "epoch": 18.251597444089455, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 22851 + }, + { + "epoch": 18.252396166134186, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22852 + }, + { + "epoch": 18.253194888178914, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 22853 + }, + { + "epoch": 18.25399361022364, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 22854 + }, + { + "epoch": 18.25479233226837, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 22855 + }, + { + "epoch": 18.2555910543131, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 22856 + }, + { + "epoch": 18.25638977635783, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 22857 + }, + { + "epoch": 18.257188498402556, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 22858 + }, + { + "epoch": 18.257987220447284, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 22859 + }, + { + "epoch": 18.25878594249201, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 22860 + }, + { + "epoch": 18.259584664536742, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 22861 + }, + { + "epoch": 18.26038338658147, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 22862 + }, + { + "epoch": 18.261182108626198, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 22863 + }, + { + "epoch": 18.261980830670925, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 22864 + }, + { + "epoch": 18.262779552715656, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 22865 + }, + { + "epoch": 18.263578274760384, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 22866 + }, + { + "epoch": 18.26437699680511, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 22867 + }, + { + "epoch": 18.26517571884984, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 22868 + }, + { + "epoch": 18.265974440894567, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 22869 + }, + { + "epoch": 18.266773162939298, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 22870 + }, + { + "epoch": 18.267571884984026, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 22871 + }, + { + "epoch": 18.268370607028753, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 22872 + }, + { + "epoch": 18.26916932907348, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 22873 + }, + { + "epoch": 18.269968051118212, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 22874 + }, + { + "epoch": 18.27076677316294, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 22875 + }, + { + "epoch": 18.271565495207668, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 22876 + }, + { + "epoch": 18.272364217252395, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 22877 + }, + { + "epoch": 18.273162939297123, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22878 + }, + { + "epoch": 18.273961661341854, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 22879 + }, + { + "epoch": 18.27476038338658, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 22880 + }, + { + "epoch": 18.27555910543131, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 22881 + }, + { + "epoch": 18.276357827476037, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 22882 + }, + { + "epoch": 18.277156549520768, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 22883 + }, + { + "epoch": 18.277955271565496, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 22884 + }, + { + "epoch": 18.278753993610223, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22885 + }, + { + "epoch": 18.27955271565495, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 22886 + }, + { + "epoch": 18.28035143769968, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 22887 + }, + { + "epoch": 18.28115015974441, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 22888 + }, + { + "epoch": 18.281948881789138, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 22889 + }, + { + "epoch": 18.282747603833865, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 22890 + }, + { + "epoch": 18.283546325878593, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22891 + }, + { + "epoch": 18.284345047923324, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 22892 + }, + { + "epoch": 18.28514376996805, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 22893 + }, + { + "epoch": 18.28594249201278, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 22894 + }, + { + "epoch": 18.286741214057507, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 22895 + }, + { + "epoch": 18.287539936102238, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 22896 + }, + { + "epoch": 18.288338658146966, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 22897 + }, + { + "epoch": 18.289137380191693, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 22898 + }, + { + "epoch": 18.28993610223642, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 22899 + }, + { + "epoch": 18.29073482428115, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 22900 + }, + { + "epoch": 18.29153354632588, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 22901 + }, + { + "epoch": 18.292332268370608, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 22902 + }, + { + "epoch": 18.293130990415335, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 22903 + }, + { + "epoch": 18.293929712460063, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 22904 + }, + { + "epoch": 18.294728434504794, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 22905 + }, + { + "epoch": 18.29552715654952, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22906 + }, + { + "epoch": 18.29632587859425, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 22907 + }, + { + "epoch": 18.297124600638977, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 22908 + }, + { + "epoch": 18.297923322683705, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 22909 + }, + { + "epoch": 18.298722044728436, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 22910 + }, + { + "epoch": 18.299520766773163, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 22911 + }, + { + "epoch": 18.30031948881789, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 22912 + }, + { + "epoch": 18.30111821086262, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22913 + }, + { + "epoch": 18.30191693290735, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 22914 + }, + { + "epoch": 18.302715654952078, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 22915 + }, + { + "epoch": 18.303514376996805, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 22916 + }, + { + "epoch": 18.304313099041533, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 22917 + }, + { + "epoch": 18.30511182108626, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 22918 + }, + { + "epoch": 18.30591054313099, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 22919 + }, + { + "epoch": 18.30670926517572, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 22920 + }, + { + "epoch": 18.307507987220447, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 22921 + }, + { + "epoch": 18.308306709265175, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 22922 + }, + { + "epoch": 18.309105431309906, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 22923 + }, + { + "epoch": 18.309904153354633, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 22924 + }, + { + "epoch": 18.31070287539936, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 22925 + }, + { + "epoch": 18.31150159744409, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 22926 + }, + { + "epoch": 18.312300319488816, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 22927 + }, + { + "epoch": 18.313099041533548, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 22928 + }, + { + "epoch": 18.313897763578275, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 22929 + }, + { + "epoch": 18.314696485623003, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 22930 + }, + { + "epoch": 18.31549520766773, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 22931 + }, + { + "epoch": 18.31629392971246, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 22932 + }, + { + "epoch": 18.31709265175719, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 22933 + }, + { + "epoch": 18.317891373801917, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 22934 + }, + { + "epoch": 18.318690095846645, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 22935 + }, + { + "epoch": 18.319488817891372, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 22936 + }, + { + "epoch": 18.320287539936103, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 22937 + }, + { + "epoch": 18.32108626198083, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 22938 + }, + { + "epoch": 18.32188498402556, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 22939 + }, + { + "epoch": 18.322683706070286, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 22940 + }, + { + "epoch": 18.323482428115017, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 22941 + }, + { + "epoch": 18.324281150159745, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 22942 + }, + { + "epoch": 18.325079872204473, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 22943 + }, + { + "epoch": 18.3258785942492, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 22944 + }, + { + "epoch": 18.326677316293928, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 22945 + }, + { + "epoch": 18.32747603833866, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 22946 + }, + { + "epoch": 18.328274760383387, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 22947 + }, + { + "epoch": 18.329073482428115, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 22948 + }, + { + "epoch": 18.329872204472842, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 22949 + }, + { + "epoch": 18.330670926517573, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 22950 + }, + { + "epoch": 18.3314696485623, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 22951 + }, + { + "epoch": 18.33226837060703, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 22952 + }, + { + "epoch": 18.333067092651756, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 22953 + }, + { + "epoch": 18.333865814696484, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 22954 + }, + { + "epoch": 18.334664536741215, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 22955 + }, + { + "epoch": 18.335463258785943, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 22956 + }, + { + "epoch": 18.33626198083067, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 22957 + }, + { + "epoch": 18.337060702875398, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 22958 + }, + { + "epoch": 18.33785942492013, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 22959 + }, + { + "epoch": 18.338658146964857, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 22960 + }, + { + "epoch": 18.339456869009584, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 22961 + }, + { + "epoch": 18.340255591054312, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 22962 + }, + { + "epoch": 18.341054313099043, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 22963 + }, + { + "epoch": 18.34185303514377, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 22964 + }, + { + "epoch": 18.3426517571885, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 22965 + }, + { + "epoch": 18.343450479233226, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22966 + }, + { + "epoch": 18.344249201277954, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 22967 + }, + { + "epoch": 18.345047923322685, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 22968 + }, + { + "epoch": 18.345846645367413, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 22969 + }, + { + "epoch": 18.34664536741214, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 22970 + }, + { + "epoch": 18.347444089456868, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 22971 + }, + { + "epoch": 18.3482428115016, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22972 + }, + { + "epoch": 18.349041533546327, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 22973 + }, + { + "epoch": 18.349840255591054, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 22974 + }, + { + "epoch": 18.350638977635782, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 22975 + }, + { + "epoch": 18.35143769968051, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 22976 + }, + { + "epoch": 18.35223642172524, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 22977 + }, + { + "epoch": 18.35303514376997, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 22978 + }, + { + "epoch": 18.353833865814696, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22979 + }, + { + "epoch": 18.354632587859424, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 22980 + }, + { + "epoch": 18.355431309904155, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 22981 + }, + { + "epoch": 18.356230031948883, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 22982 + }, + { + "epoch": 18.35702875399361, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 22983 + }, + { + "epoch": 18.357827476038338, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 22984 + }, + { + "epoch": 18.358626198083066, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 22985 + }, + { + "epoch": 18.359424920127797, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 22986 + }, + { + "epoch": 18.360223642172524, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 22987 + }, + { + "epoch": 18.361022364217252, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 22988 + }, + { + "epoch": 18.36182108626198, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 22989 + }, + { + "epoch": 18.36261980830671, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 22990 + }, + { + "epoch": 18.36341853035144, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 22991 + }, + { + "epoch": 18.364217252396166, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 22992 + }, + { + "epoch": 18.365015974440894, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 22993 + }, + { + "epoch": 18.36581469648562, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 22994 + }, + { + "epoch": 18.366613418530353, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 22995 + }, + { + "epoch": 18.36741214057508, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 22996 + }, + { + "epoch": 18.368210862619808, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 22997 + }, + { + "epoch": 18.369009584664536, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0701, + "step": 22998 + }, + { + "epoch": 18.369808306709267, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 22999 + }, + { + "epoch": 18.370607028753994, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23000 + }, + { + "epoch": 18.371405750798722, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23001 + }, + { + "epoch": 18.37220447284345, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 23002 + }, + { + "epoch": 18.373003194888177, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 23003 + }, + { + "epoch": 18.37380191693291, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 23004 + }, + { + "epoch": 18.374600638977636, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0753, + "step": 23005 + }, + { + "epoch": 18.375399361022364, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 23006 + }, + { + "epoch": 18.37619808306709, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 23007 + }, + { + "epoch": 18.376996805111823, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 23008 + }, + { + "epoch": 18.37779552715655, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 23009 + }, + { + "epoch": 18.378594249201278, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23010 + }, + { + "epoch": 18.379392971246006, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 23011 + }, + { + "epoch": 18.380191693290733, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23012 + }, + { + "epoch": 18.380990415335464, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 23013 + }, + { + "epoch": 18.381789137380192, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 23014 + }, + { + "epoch": 18.38258785942492, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23015 + }, + { + "epoch": 18.383386581469647, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 23016 + }, + { + "epoch": 18.38418530351438, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 23017 + }, + { + "epoch": 18.384984025559106, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 23018 + }, + { + "epoch": 18.385782747603834, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23019 + }, + { + "epoch": 18.38658146964856, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 23020 + }, + { + "epoch": 18.38738019169329, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 23021 + }, + { + "epoch": 18.38817891373802, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23022 + }, + { + "epoch": 18.388977635782748, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 23023 + }, + { + "epoch": 18.389776357827476, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 23024 + }, + { + "epoch": 18.390575079872203, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 23025 + }, + { + "epoch": 18.391373801916934, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 23026 + }, + { + "epoch": 18.392172523961662, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 23027 + }, + { + "epoch": 18.39297124600639, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 23028 + }, + { + "epoch": 18.393769968051117, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 23029 + }, + { + "epoch": 18.394568690095845, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 23030 + }, + { + "epoch": 18.395367412140576, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 23031 + }, + { + "epoch": 18.396166134185304, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 23032 + }, + { + "epoch": 18.39696485623003, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23033 + }, + { + "epoch": 18.39776357827476, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 23034 + }, + { + "epoch": 18.39856230031949, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 23035 + }, + { + "epoch": 18.399361022364218, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23036 + }, + { + "epoch": 18.400159744408946, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 23037 + }, + { + "epoch": 18.400958466453673, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23038 + }, + { + "epoch": 18.4017571884984, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 23039 + }, + { + "epoch": 18.402555910543132, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 23040 + }, + { + "epoch": 18.40335463258786, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 23041 + }, + { + "epoch": 18.404153354632587, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 23042 + }, + { + "epoch": 18.404952076677315, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 23043 + }, + { + "epoch": 18.405750798722046, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 23044 + }, + { + "epoch": 18.406549520766774, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 23045 + }, + { + "epoch": 18.4073482428115, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 23046 + }, + { + "epoch": 18.40814696485623, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 23047 + }, + { + "epoch": 18.408945686900957, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 23048 + }, + { + "epoch": 18.409744408945688, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 23049 + }, + { + "epoch": 18.410543130990416, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 23050 + }, + { + "epoch": 18.411341853035143, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 23051 + }, + { + "epoch": 18.41214057507987, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 23052 + }, + { + "epoch": 18.412939297124602, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 23053 + }, + { + "epoch": 18.41373801916933, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 23054 + }, + { + "epoch": 18.414536741214057, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 23055 + }, + { + "epoch": 18.415335463258785, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 23056 + }, + { + "epoch": 18.416134185303516, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 23057 + }, + { + "epoch": 18.416932907348244, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 23058 + }, + { + "epoch": 18.41773162939297, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 23059 + }, + { + "epoch": 18.4185303514377, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 23060 + }, + { + "epoch": 18.419329073482427, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 23061 + }, + { + "epoch": 18.420127795527158, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 23062 + }, + { + "epoch": 18.420926517571885, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 23063 + }, + { + "epoch": 18.421725239616613, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 23064 + }, + { + "epoch": 18.42252396166134, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 23065 + }, + { + "epoch": 18.423322683706072, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 23066 + }, + { + "epoch": 18.4241214057508, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23067 + }, + { + "epoch": 18.424920127795527, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 23068 + }, + { + "epoch": 18.425718849840255, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 23069 + }, + { + "epoch": 18.426517571884983, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 23070 + }, + { + "epoch": 18.427316293929714, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 23071 + }, + { + "epoch": 18.42811501597444, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 23072 + }, + { + "epoch": 18.42891373801917, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 23073 + }, + { + "epoch": 18.429712460063897, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 23074 + }, + { + "epoch": 18.430511182108628, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 23075 + }, + { + "epoch": 18.431309904153355, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 23076 + }, + { + "epoch": 18.432108626198083, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 23077 + }, + { + "epoch": 18.43290734824281, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23078 + }, + { + "epoch": 18.43370607028754, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 23079 + }, + { + "epoch": 18.43450479233227, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 23080 + }, + { + "epoch": 18.435303514376997, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 23081 + }, + { + "epoch": 18.436102236421725, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 23082 + }, + { + "epoch": 18.436900958466452, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 23083 + }, + { + "epoch": 18.437699680511184, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 23084 + }, + { + "epoch": 18.43849840255591, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23085 + }, + { + "epoch": 18.43929712460064, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 23086 + }, + { + "epoch": 18.440095846645367, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 23087 + }, + { + "epoch": 18.440894568690094, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 23088 + }, + { + "epoch": 18.441693290734825, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 23089 + }, + { + "epoch": 18.442492012779553, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 23090 + }, + { + "epoch": 18.44329073482428, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 23091 + }, + { + "epoch": 18.44408945686901, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 23092 + }, + { + "epoch": 18.44488817891374, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 23093 + }, + { + "epoch": 18.445686900958467, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 23094 + }, + { + "epoch": 18.446485623003195, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 23095 + }, + { + "epoch": 18.447284345047922, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 23096 + }, + { + "epoch": 18.44808306709265, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 23097 + }, + { + "epoch": 18.44888178913738, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 23098 + }, + { + "epoch": 18.44968051118211, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 23099 + }, + { + "epoch": 18.450479233226837, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 23100 + }, + { + "epoch": 18.451277955271564, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 23101 + }, + { + "epoch": 18.452076677316295, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23102 + }, + { + "epoch": 18.452875399361023, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 23103 + }, + { + "epoch": 18.45367412140575, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 23104 + }, + { + "epoch": 18.45447284345048, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 23105 + }, + { + "epoch": 18.455271565495206, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 23106 + }, + { + "epoch": 18.456070287539937, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 23107 + }, + { + "epoch": 18.456869009584665, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 23108 + }, + { + "epoch": 18.457667731629392, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 23109 + }, + { + "epoch": 18.45846645367412, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 23110 + }, + { + "epoch": 18.45926517571885, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23111 + }, + { + "epoch": 18.46006389776358, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 23112 + }, + { + "epoch": 18.460862619808307, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 23113 + }, + { + "epoch": 18.461661341853034, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 23114 + }, + { + "epoch": 18.462460063897762, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 23115 + }, + { + "epoch": 18.463258785942493, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23116 + }, + { + "epoch": 18.46405750798722, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 23117 + }, + { + "epoch": 18.46485623003195, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 23118 + }, + { + "epoch": 18.465654952076676, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 23119 + }, + { + "epoch": 18.466453674121407, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 23120 + }, + { + "epoch": 18.467252396166135, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 23121 + }, + { + "epoch": 18.468051118210862, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 23122 + }, + { + "epoch": 18.46884984025559, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 23123 + }, + { + "epoch": 18.46964856230032, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 23124 + }, + { + "epoch": 18.47044728434505, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 23125 + }, + { + "epoch": 18.471246006389777, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 23126 + }, + { + "epoch": 18.472044728434504, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 23127 + }, + { + "epoch": 18.472843450479232, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 23128 + }, + { + "epoch": 18.473642172523963, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 23129 + }, + { + "epoch": 18.47444089456869, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 23130 + }, + { + "epoch": 18.47523961661342, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 23131 + }, + { + "epoch": 18.476038338658146, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 23132 + }, + { + "epoch": 18.476837060702877, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 23133 + }, + { + "epoch": 18.477635782747605, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 23134 + }, + { + "epoch": 18.478434504792332, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 23135 + }, + { + "epoch": 18.47923322683706, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23136 + }, + { + "epoch": 18.480031948881788, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23137 + }, + { + "epoch": 18.48083067092652, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 23138 + }, + { + "epoch": 18.481629392971247, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 23139 + }, + { + "epoch": 18.482428115015974, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 23140 + }, + { + "epoch": 18.483226837060702, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 23141 + }, + { + "epoch": 18.484025559105433, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 23142 + }, + { + "epoch": 18.48482428115016, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 23143 + }, + { + "epoch": 18.48562300319489, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 23144 + }, + { + "epoch": 18.486421725239616, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 23145 + }, + { + "epoch": 18.487220447284344, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 23146 + }, + { + "epoch": 18.488019169329075, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 23147 + }, + { + "epoch": 18.488817891373802, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 23148 + }, + { + "epoch": 18.48961661341853, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 23149 + }, + { + "epoch": 18.490415335463258, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23150 + }, + { + "epoch": 18.49121405750799, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 23151 + }, + { + "epoch": 18.492012779552716, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 23152 + }, + { + "epoch": 18.492811501597444, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23153 + }, + { + "epoch": 18.49361022364217, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 23154 + }, + { + "epoch": 18.4944089456869, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 23155 + }, + { + "epoch": 18.49520766773163, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 23156 + }, + { + "epoch": 18.49600638977636, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 23157 + }, + { + "epoch": 18.496805111821086, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23158 + }, + { + "epoch": 18.497603833865814, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 23159 + }, + { + "epoch": 18.498402555910545, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 23160 + }, + { + "epoch": 18.499201277955272, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 23161 + }, + { + "epoch": 18.5, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23162 + }, + { + "epoch": 18.500798722044728, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 23163 + }, + { + "epoch": 18.501597444089455, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 23164 + }, + { + "epoch": 18.502396166134186, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 23165 + }, + { + "epoch": 18.503194888178914, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 23166 + }, + { + "epoch": 18.50399361022364, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23167 + }, + { + "epoch": 18.50479233226837, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23168 + }, + { + "epoch": 18.5055910543131, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23169 + }, + { + "epoch": 18.50638977635783, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 23170 + }, + { + "epoch": 18.507188498402556, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 23171 + }, + { + "epoch": 18.507987220447284, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 23172 + }, + { + "epoch": 18.50878594249201, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 23173 + }, + { + "epoch": 18.509584664536742, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 23174 + }, + { + "epoch": 18.51038338658147, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 23175 + }, + { + "epoch": 18.511182108626198, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 23176 + }, + { + "epoch": 18.511980830670925, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 23177 + }, + { + "epoch": 18.512779552715656, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 23178 + }, + { + "epoch": 18.513578274760384, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 23179 + }, + { + "epoch": 18.51437699680511, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 23180 + }, + { + "epoch": 18.51517571884984, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 23181 + }, + { + "epoch": 18.515974440894567, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23182 + }, + { + "epoch": 18.516773162939298, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 23183 + }, + { + "epoch": 18.517571884984026, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 23184 + }, + { + "epoch": 18.518370607028753, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 23185 + }, + { + "epoch": 18.51916932907348, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 23186 + }, + { + "epoch": 18.519968051118212, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 23187 + }, + { + "epoch": 18.52076677316294, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 23188 + }, + { + "epoch": 18.521565495207668, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 23189 + }, + { + "epoch": 18.522364217252395, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 23190 + }, + { + "epoch": 18.523162939297123, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 23191 + }, + { + "epoch": 18.523961661341854, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 23192 + }, + { + "epoch": 18.52476038338658, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 23193 + }, + { + "epoch": 18.52555910543131, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 23194 + }, + { + "epoch": 18.526357827476037, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23195 + }, + { + "epoch": 18.527156549520768, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 23196 + }, + { + "epoch": 18.527955271565496, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 23197 + }, + { + "epoch": 18.528753993610223, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 23198 + }, + { + "epoch": 18.52955271565495, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 23199 + }, + { + "epoch": 18.53035143769968, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 23200 + }, + { + "epoch": 18.53115015974441, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 23201 + }, + { + "epoch": 18.531948881789138, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 23202 + }, + { + "epoch": 18.532747603833865, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 23203 + }, + { + "epoch": 18.533546325878593, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 23204 + }, + { + "epoch": 18.534345047923324, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 23205 + }, + { + "epoch": 18.53514376996805, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 23206 + }, + { + "epoch": 18.53594249201278, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 23207 + }, + { + "epoch": 18.536741214057507, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 23208 + }, + { + "epoch": 18.537539936102235, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 23209 + }, + { + "epoch": 18.538338658146966, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 23210 + }, + { + "epoch": 18.539137380191693, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 23211 + }, + { + "epoch": 18.53993610223642, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 23212 + }, + { + "epoch": 18.54073482428115, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 23213 + }, + { + "epoch": 18.54153354632588, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23214 + }, + { + "epoch": 18.542332268370608, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 23215 + }, + { + "epoch": 18.543130990415335, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 23216 + }, + { + "epoch": 18.543929712460063, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 23217 + }, + { + "epoch": 18.544728434504794, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 23218 + }, + { + "epoch": 18.54552715654952, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 23219 + }, + { + "epoch": 18.54632587859425, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 23220 + }, + { + "epoch": 18.547124600638977, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 23221 + }, + { + "epoch": 18.547923322683705, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 23222 + }, + { + "epoch": 18.548722044728436, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 23223 + }, + { + "epoch": 18.549520766773163, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 23224 + }, + { + "epoch": 18.55031948881789, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 23225 + }, + { + "epoch": 18.55111821086262, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23226 + }, + { + "epoch": 18.55191693290735, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 23227 + }, + { + "epoch": 18.552715654952078, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 23228 + }, + { + "epoch": 18.553514376996805, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 23229 + }, + { + "epoch": 18.554313099041533, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 23230 + }, + { + "epoch": 18.55511182108626, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 23231 + }, + { + "epoch": 18.55591054313099, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 23232 + }, + { + "epoch": 18.55670926517572, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 23233 + }, + { + "epoch": 18.557507987220447, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 23234 + }, + { + "epoch": 18.558306709265175, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 23235 + }, + { + "epoch": 18.559105431309906, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 23236 + }, + { + "epoch": 18.559904153354633, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 23237 + }, + { + "epoch": 18.56070287539936, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23238 + }, + { + "epoch": 18.56150159744409, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 23239 + }, + { + "epoch": 18.562300319488816, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 23240 + }, + { + "epoch": 18.563099041533548, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 23241 + }, + { + "epoch": 18.563897763578275, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 23242 + }, + { + "epoch": 18.564696485623003, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 23243 + }, + { + "epoch": 18.56549520766773, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 23244 + }, + { + "epoch": 18.56629392971246, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 23245 + }, + { + "epoch": 18.56709265175719, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 23246 + }, + { + "epoch": 18.567891373801917, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 23247 + }, + { + "epoch": 18.568690095846645, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23248 + }, + { + "epoch": 18.569488817891372, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 23249 + }, + { + "epoch": 18.570287539936103, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 23250 + }, + { + "epoch": 18.57108626198083, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23251 + }, + { + "epoch": 18.57188498402556, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 23252 + }, + { + "epoch": 18.572683706070286, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 23253 + }, + { + "epoch": 18.573482428115017, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 23254 + }, + { + "epoch": 18.574281150159745, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 23255 + }, + { + "epoch": 18.575079872204473, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 23256 + }, + { + "epoch": 18.5758785942492, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 23257 + }, + { + "epoch": 18.576677316293928, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 23258 + }, + { + "epoch": 18.57747603833866, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 23259 + }, + { + "epoch": 18.578274760383387, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 23260 + }, + { + "epoch": 18.579073482428115, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 23261 + }, + { + "epoch": 18.579872204472842, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 23262 + }, + { + "epoch": 18.580670926517573, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 23263 + }, + { + "epoch": 18.5814696485623, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 23264 + }, + { + "epoch": 18.58226837060703, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 23265 + }, + { + "epoch": 18.583067092651756, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23266 + }, + { + "epoch": 18.583865814696484, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 23267 + }, + { + "epoch": 18.584664536741215, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 23268 + }, + { + "epoch": 18.585463258785943, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 23269 + }, + { + "epoch": 18.58626198083067, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 23270 + }, + { + "epoch": 18.587060702875398, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 23271 + }, + { + "epoch": 18.58785942492013, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 23272 + }, + { + "epoch": 18.588658146964857, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 23273 + }, + { + "epoch": 18.589456869009584, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 23274 + }, + { + "epoch": 18.590255591054312, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 23275 + }, + { + "epoch": 18.591054313099043, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 23276 + }, + { + "epoch": 18.59185303514377, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 23277 + }, + { + "epoch": 18.5926517571885, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 23278 + }, + { + "epoch": 18.593450479233226, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 23279 + }, + { + "epoch": 18.594249201277954, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 23280 + }, + { + "epoch": 18.595047923322685, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 23281 + }, + { + "epoch": 18.595846645367413, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 23282 + }, + { + "epoch": 18.59664536741214, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 23283 + }, + { + "epoch": 18.597444089456868, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 23284 + }, + { + "epoch": 18.5982428115016, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 23285 + }, + { + "epoch": 18.599041533546327, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 23286 + }, + { + "epoch": 18.599840255591054, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 23287 + }, + { + "epoch": 18.600638977635782, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 23288 + }, + { + "epoch": 18.60143769968051, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 23289 + }, + { + "epoch": 18.60223642172524, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 23290 + }, + { + "epoch": 18.60303514376997, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 23291 + }, + { + "epoch": 18.603833865814696, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 23292 + }, + { + "epoch": 18.604632587859424, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23293 + }, + { + "epoch": 18.605431309904155, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 23294 + }, + { + "epoch": 18.606230031948883, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 23295 + }, + { + "epoch": 18.60702875399361, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 23296 + }, + { + "epoch": 18.607827476038338, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 23297 + }, + { + "epoch": 18.608626198083066, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 23298 + }, + { + "epoch": 18.609424920127797, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 23299 + }, + { + "epoch": 18.610223642172524, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23300 + }, + { + "epoch": 18.611022364217252, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23301 + }, + { + "epoch": 18.61182108626198, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 23302 + }, + { + "epoch": 18.61261980830671, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 23303 + }, + { + "epoch": 18.61341853035144, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 23304 + }, + { + "epoch": 18.614217252396166, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 23305 + }, + { + "epoch": 18.615015974440894, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 23306 + }, + { + "epoch": 18.61581469648562, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 23307 + }, + { + "epoch": 18.616613418530353, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 23308 + }, + { + "epoch": 18.61741214057508, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 23309 + }, + { + "epoch": 18.618210862619808, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 23310 + }, + { + "epoch": 18.619009584664536, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 23311 + }, + { + "epoch": 18.619808306709267, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1097, + "step": 23312 + }, + { + "epoch": 18.620607028753994, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 23313 + }, + { + "epoch": 18.621405750798722, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 23314 + }, + { + "epoch": 18.62220447284345, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 23315 + }, + { + "epoch": 18.623003194888177, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 23316 + }, + { + "epoch": 18.62380191693291, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 23317 + }, + { + "epoch": 18.624600638977636, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 23318 + }, + { + "epoch": 18.625399361022364, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 23319 + }, + { + "epoch": 18.62619808306709, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 23320 + }, + { + "epoch": 18.626996805111823, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23321 + }, + { + "epoch": 18.62779552715655, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 23322 + }, + { + "epoch": 18.628594249201278, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 23323 + }, + { + "epoch": 18.629392971246006, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 23324 + }, + { + "epoch": 18.630191693290733, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 23325 + }, + { + "epoch": 18.630990415335464, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23326 + }, + { + "epoch": 18.631789137380192, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 23327 + }, + { + "epoch": 18.63258785942492, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 23328 + }, + { + "epoch": 18.633386581469647, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23329 + }, + { + "epoch": 18.63418530351438, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 23330 + }, + { + "epoch": 18.634984025559106, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 23331 + }, + { + "epoch": 18.635782747603834, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 23332 + }, + { + "epoch": 18.63658146964856, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 23333 + }, + { + "epoch": 18.63738019169329, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 23334 + }, + { + "epoch": 18.63817891373802, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 23335 + }, + { + "epoch": 18.638977635782748, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 23336 + }, + { + "epoch": 18.639776357827476, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 23337 + }, + { + "epoch": 18.640575079872203, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 23338 + }, + { + "epoch": 18.641373801916934, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 23339 + }, + { + "epoch": 18.642172523961662, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 23340 + }, + { + "epoch": 18.64297124600639, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 23341 + }, + { + "epoch": 18.643769968051117, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 23342 + }, + { + "epoch": 18.644568690095845, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23343 + }, + { + "epoch": 18.645367412140576, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 23344 + }, + { + "epoch": 18.646166134185304, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23345 + }, + { + "epoch": 18.64696485623003, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 23346 + }, + { + "epoch": 18.64776357827476, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 23347 + }, + { + "epoch": 18.64856230031949, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 23348 + }, + { + "epoch": 18.649361022364218, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 23349 + }, + { + "epoch": 18.650159744408946, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 23350 + }, + { + "epoch": 18.650958466453673, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23351 + }, + { + "epoch": 18.6517571884984, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23352 + }, + { + "epoch": 18.652555910543132, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 23353 + }, + { + "epoch": 18.65335463258786, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 23354 + }, + { + "epoch": 18.654153354632587, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 23355 + }, + { + "epoch": 18.654952076677315, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1091, + "step": 23356 + }, + { + "epoch": 18.655750798722046, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 23357 + }, + { + "epoch": 18.656549520766774, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 23358 + }, + { + "epoch": 18.6573482428115, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 23359 + }, + { + "epoch": 18.65814696485623, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23360 + }, + { + "epoch": 18.658945686900957, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 23361 + }, + { + "epoch": 18.659744408945688, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 23362 + }, + { + "epoch": 18.660543130990416, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23363 + }, + { + "epoch": 18.661341853035143, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23364 + }, + { + "epoch": 18.66214057507987, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 23365 + }, + { + "epoch": 18.662939297124602, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 23366 + }, + { + "epoch": 18.66373801916933, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 23367 + }, + { + "epoch": 18.664536741214057, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 23368 + }, + { + "epoch": 18.665335463258785, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 23369 + }, + { + "epoch": 18.666134185303516, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 23370 + }, + { + "epoch": 18.666932907348244, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 23371 + }, + { + "epoch": 18.66773162939297, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 23372 + }, + { + "epoch": 18.6685303514377, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 23373 + }, + { + "epoch": 18.669329073482427, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 23374 + }, + { + "epoch": 18.670127795527158, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 23375 + }, + { + "epoch": 18.670926517571885, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 23376 + }, + { + "epoch": 18.671725239616613, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23377 + }, + { + "epoch": 18.67252396166134, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 23378 + }, + { + "epoch": 18.673322683706072, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 23379 + }, + { + "epoch": 18.6741214057508, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23380 + }, + { + "epoch": 18.674920127795527, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 23381 + }, + { + "epoch": 18.675718849840255, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23382 + }, + { + "epoch": 18.676517571884983, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 23383 + }, + { + "epoch": 18.677316293929714, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 23384 + }, + { + "epoch": 18.67811501597444, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23385 + }, + { + "epoch": 18.67891373801917, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 23386 + }, + { + "epoch": 18.679712460063897, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 23387 + }, + { + "epoch": 18.680511182108628, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 23388 + }, + { + "epoch": 18.681309904153355, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 23389 + }, + { + "epoch": 18.682108626198083, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 23390 + }, + { + "epoch": 18.68290734824281, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 23391 + }, + { + "epoch": 18.68370607028754, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23392 + }, + { + "epoch": 18.68450479233227, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 23393 + }, + { + "epoch": 18.685303514376997, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 23394 + }, + { + "epoch": 18.686102236421725, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23395 + }, + { + "epoch": 18.686900958466452, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 23396 + }, + { + "epoch": 18.687699680511184, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 23397 + }, + { + "epoch": 18.68849840255591, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 23398 + }, + { + "epoch": 18.68929712460064, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 23399 + }, + { + "epoch": 18.690095846645367, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 23400 + }, + { + "epoch": 18.690894568690094, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 23401 + }, + { + "epoch": 18.691693290734825, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 23402 + }, + { + "epoch": 18.692492012779553, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 23403 + }, + { + "epoch": 18.69329073482428, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 23404 + }, + { + "epoch": 18.69408945686901, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23405 + }, + { + "epoch": 18.69488817891374, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23406 + }, + { + "epoch": 18.695686900958467, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 23407 + }, + { + "epoch": 18.696485623003195, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23408 + }, + { + "epoch": 18.697284345047922, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23409 + }, + { + "epoch": 18.69808306709265, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 23410 + }, + { + "epoch": 18.69888178913738, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1038, + "step": 23411 + }, + { + "epoch": 18.69968051118211, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 23412 + }, + { + "epoch": 18.700479233226837, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 23413 + }, + { + "epoch": 18.701277955271564, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 23414 + }, + { + "epoch": 18.702076677316295, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 23415 + }, + { + "epoch": 18.702875399361023, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 23416 + }, + { + "epoch": 18.70367412140575, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23417 + }, + { + "epoch": 18.70447284345048, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 23418 + }, + { + "epoch": 18.705271565495206, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 23419 + }, + { + "epoch": 18.706070287539937, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 23420 + }, + { + "epoch": 18.706869009584665, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 23421 + }, + { + "epoch": 18.707667731629392, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 23422 + }, + { + "epoch": 18.70846645367412, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23423 + }, + { + "epoch": 18.70926517571885, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 23424 + }, + { + "epoch": 18.71006389776358, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 23425 + }, + { + "epoch": 18.710862619808307, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 23426 + }, + { + "epoch": 18.711661341853034, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 23427 + }, + { + "epoch": 18.712460063897765, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 23428 + }, + { + "epoch": 18.713258785942493, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 23429 + }, + { + "epoch": 18.71405750798722, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 23430 + }, + { + "epoch": 18.71485623003195, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 23431 + }, + { + "epoch": 18.715654952076676, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 23432 + }, + { + "epoch": 18.716453674121407, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 23433 + }, + { + "epoch": 18.717252396166135, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23434 + }, + { + "epoch": 18.718051118210862, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 23435 + }, + { + "epoch": 18.71884984025559, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 23436 + }, + { + "epoch": 18.71964856230032, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 23437 + }, + { + "epoch": 18.72044728434505, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 23438 + }, + { + "epoch": 18.721246006389777, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 23439 + }, + { + "epoch": 18.722044728434504, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23440 + }, + { + "epoch": 18.722843450479232, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 23441 + }, + { + "epoch": 18.723642172523963, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 23442 + }, + { + "epoch": 18.72444089456869, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 23443 + }, + { + "epoch": 18.72523961661342, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23444 + }, + { + "epoch": 18.726038338658146, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 23445 + }, + { + "epoch": 18.726837060702877, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 23446 + }, + { + "epoch": 18.727635782747605, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 23447 + }, + { + "epoch": 18.728434504792332, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 23448 + }, + { + "epoch": 18.72923322683706, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 23449 + }, + { + "epoch": 18.730031948881788, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 23450 + }, + { + "epoch": 18.73083067092652, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 23451 + }, + { + "epoch": 18.731629392971247, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 23452 + }, + { + "epoch": 18.732428115015974, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 23453 + }, + { + "epoch": 18.733226837060702, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 23454 + }, + { + "epoch": 18.734025559105433, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 23455 + }, + { + "epoch": 18.73482428115016, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 23456 + }, + { + "epoch": 18.73562300319489, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 23457 + }, + { + "epoch": 18.736421725239616, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 23458 + }, + { + "epoch": 18.737220447284344, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 23459 + }, + { + "epoch": 18.738019169329075, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 23460 + }, + { + "epoch": 18.738817891373802, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 23461 + }, + { + "epoch": 18.73961661341853, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 23462 + }, + { + "epoch": 18.740415335463258, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 23463 + }, + { + "epoch": 18.74121405750799, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 23464 + }, + { + "epoch": 18.742012779552716, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 23465 + }, + { + "epoch": 18.742811501597444, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 23466 + }, + { + "epoch": 18.74361022364217, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 23467 + }, + { + "epoch": 18.7444089456869, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 23468 + }, + { + "epoch": 18.74520766773163, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 23469 + }, + { + "epoch": 18.74600638977636, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 23470 + }, + { + "epoch": 18.746805111821086, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 23471 + }, + { + "epoch": 18.747603833865814, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 23472 + }, + { + "epoch": 18.748402555910545, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 23473 + }, + { + "epoch": 18.749201277955272, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 23474 + }, + { + "epoch": 18.75, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 23475 + }, + { + "epoch": 18.750798722044728, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 23476 + }, + { + "epoch": 18.751597444089455, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 23477 + }, + { + "epoch": 18.752396166134186, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 23478 + }, + { + "epoch": 18.753194888178914, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 23479 + }, + { + "epoch": 18.75399361022364, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 23480 + }, + { + "epoch": 18.75479233226837, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23481 + }, + { + "epoch": 18.7555910543131, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 23482 + }, + { + "epoch": 18.75638977635783, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23483 + }, + { + "epoch": 18.757188498402556, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23484 + }, + { + "epoch": 18.757987220447284, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 23485 + }, + { + "epoch": 18.75878594249201, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 23486 + }, + { + "epoch": 18.759584664536742, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 23487 + }, + { + "epoch": 18.76038338658147, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 23488 + }, + { + "epoch": 18.761182108626198, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 23489 + }, + { + "epoch": 18.761980830670925, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 23490 + }, + { + "epoch": 18.762779552715656, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 23491 + }, + { + "epoch": 18.763578274760384, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 23492 + }, + { + "epoch": 18.76437699680511, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23493 + }, + { + "epoch": 18.76517571884984, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 23494 + }, + { + "epoch": 18.765974440894567, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 23495 + }, + { + "epoch": 18.766773162939298, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 23496 + }, + { + "epoch": 18.767571884984026, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 23497 + }, + { + "epoch": 18.768370607028753, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 23498 + }, + { + "epoch": 18.76916932907348, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 23499 + }, + { + "epoch": 18.769968051118212, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 23500 + }, + { + "epoch": 18.77076677316294, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23501 + }, + { + "epoch": 18.771565495207668, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23502 + }, + { + "epoch": 18.772364217252395, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23503 + }, + { + "epoch": 18.773162939297123, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 23504 + }, + { + "epoch": 18.773961661341854, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 23505 + }, + { + "epoch": 18.77476038338658, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23506 + }, + { + "epoch": 18.77555910543131, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 23507 + }, + { + "epoch": 18.776357827476037, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 23508 + }, + { + "epoch": 18.777156549520768, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 23509 + }, + { + "epoch": 18.777955271565496, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 23510 + }, + { + "epoch": 18.778753993610223, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1067, + "step": 23511 + }, + { + "epoch": 18.77955271565495, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 23512 + }, + { + "epoch": 18.78035143769968, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 23513 + }, + { + "epoch": 18.78115015974441, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 23514 + }, + { + "epoch": 18.781948881789138, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 23515 + }, + { + "epoch": 18.782747603833865, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 23516 + }, + { + "epoch": 18.783546325878593, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 23517 + }, + { + "epoch": 18.784345047923324, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 23518 + }, + { + "epoch": 18.78514376996805, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 23519 + }, + { + "epoch": 18.78594249201278, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 23520 + }, + { + "epoch": 18.786741214057507, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 23521 + }, + { + "epoch": 18.787539936102235, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 23522 + }, + { + "epoch": 18.788338658146966, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 23523 + }, + { + "epoch": 18.789137380191693, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 23524 + }, + { + "epoch": 18.78993610223642, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 23525 + }, + { + "epoch": 18.79073482428115, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 23526 + }, + { + "epoch": 18.79153354632588, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 23527 + }, + { + "epoch": 18.792332268370608, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 23528 + }, + { + "epoch": 18.793130990415335, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 23529 + }, + { + "epoch": 18.793929712460063, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 23530 + }, + { + "epoch": 18.794728434504794, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 23531 + }, + { + "epoch": 18.79552715654952, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 23532 + }, + { + "epoch": 18.79632587859425, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 23533 + }, + { + "epoch": 18.797124600638977, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 23534 + }, + { + "epoch": 18.797923322683705, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 23535 + }, + { + "epoch": 18.798722044728436, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 23536 + }, + { + "epoch": 18.799520766773163, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 23537 + }, + { + "epoch": 18.80031948881789, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 23538 + }, + { + "epoch": 18.80111821086262, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 23539 + }, + { + "epoch": 18.80191693290735, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 23540 + }, + { + "epoch": 18.802715654952078, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 23541 + }, + { + "epoch": 18.803514376996805, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 23542 + }, + { + "epoch": 18.804313099041533, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 23543 + }, + { + "epoch": 18.80511182108626, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23544 + }, + { + "epoch": 18.80591054313099, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 23545 + }, + { + "epoch": 18.80670926517572, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 23546 + }, + { + "epoch": 18.807507987220447, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23547 + }, + { + "epoch": 18.808306709265175, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 23548 + }, + { + "epoch": 18.809105431309906, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 23549 + }, + { + "epoch": 18.809904153354633, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1072, + "step": 23550 + }, + { + "epoch": 18.81070287539936, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1089, + "step": 23551 + }, + { + "epoch": 18.81150159744409, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 23552 + }, + { + "epoch": 18.812300319488816, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 23553 + }, + { + "epoch": 18.813099041533548, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 23554 + }, + { + "epoch": 18.813897763578275, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 23555 + }, + { + "epoch": 18.814696485623003, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 23556 + }, + { + "epoch": 18.81549520766773, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 23557 + }, + { + "epoch": 18.81629392971246, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23558 + }, + { + "epoch": 18.81709265175719, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 23559 + }, + { + "epoch": 18.817891373801917, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 23560 + }, + { + "epoch": 18.818690095846645, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 23561 + }, + { + "epoch": 18.819488817891372, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 23562 + }, + { + "epoch": 18.820287539936103, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 23563 + }, + { + "epoch": 18.82108626198083, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23564 + }, + { + "epoch": 18.82188498402556, + "grad_norm": 0.3671875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 23565 + }, + { + "epoch": 18.822683706070286, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 23566 + }, + { + "epoch": 18.823482428115017, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 23567 + }, + { + "epoch": 18.824281150159745, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 23568 + }, + { + "epoch": 18.825079872204473, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 23569 + }, + { + "epoch": 18.8258785942492, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23570 + }, + { + "epoch": 18.826677316293928, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 23571 + }, + { + "epoch": 18.82747603833866, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 23572 + }, + { + "epoch": 18.828274760383387, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 23573 + }, + { + "epoch": 18.829073482428115, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 23574 + }, + { + "epoch": 18.829872204472842, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 23575 + }, + { + "epoch": 18.830670926517573, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 23576 + }, + { + "epoch": 18.8314696485623, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 23577 + }, + { + "epoch": 18.83226837060703, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 23578 + }, + { + "epoch": 18.833067092651756, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 23579 + }, + { + "epoch": 18.833865814696484, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 23580 + }, + { + "epoch": 18.834664536741215, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 23581 + }, + { + "epoch": 18.835463258785943, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 23582 + }, + { + "epoch": 18.83626198083067, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 23583 + }, + { + "epoch": 18.837060702875398, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 23584 + }, + { + "epoch": 18.83785942492013, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23585 + }, + { + "epoch": 18.838658146964857, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 23586 + }, + { + "epoch": 18.839456869009584, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 23587 + }, + { + "epoch": 18.840255591054312, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 23588 + }, + { + "epoch": 18.841054313099043, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23589 + }, + { + "epoch": 18.84185303514377, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 23590 + }, + { + "epoch": 18.8426517571885, + "grad_norm": 0.095703125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 23591 + }, + { + "epoch": 18.843450479233226, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 23592 + }, + { + "epoch": 18.844249201277954, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 23593 + }, + { + "epoch": 18.845047923322685, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 23594 + }, + { + "epoch": 18.845846645367413, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 23595 + }, + { + "epoch": 18.84664536741214, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 23596 + }, + { + "epoch": 18.847444089456868, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 23597 + }, + { + "epoch": 18.8482428115016, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 23598 + }, + { + "epoch": 18.849041533546327, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 23599 + }, + { + "epoch": 18.849840255591054, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 23600 + }, + { + "epoch": 18.850638977635782, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 23601 + }, + { + "epoch": 18.85143769968051, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 23602 + }, + { + "epoch": 18.85223642172524, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 23603 + }, + { + "epoch": 18.85303514376997, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 23604 + }, + { + "epoch": 18.853833865814696, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 23605 + }, + { + "epoch": 18.854632587859424, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 23606 + }, + { + "epoch": 18.855431309904155, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 23607 + }, + { + "epoch": 18.856230031948883, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 23608 + }, + { + "epoch": 18.85702875399361, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 23609 + }, + { + "epoch": 18.857827476038338, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 23610 + }, + { + "epoch": 18.858626198083066, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 23611 + }, + { + "epoch": 18.859424920127797, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 23612 + }, + { + "epoch": 18.860223642172524, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 23613 + }, + { + "epoch": 18.861022364217252, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 23614 + }, + { + "epoch": 18.86182108626198, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 23615 + }, + { + "epoch": 18.86261980830671, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 23616 + }, + { + "epoch": 18.86341853035144, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 23617 + }, + { + "epoch": 18.864217252396166, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 23618 + }, + { + "epoch": 18.865015974440894, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 23619 + }, + { + "epoch": 18.86581469648562, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 23620 + }, + { + "epoch": 18.866613418530353, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 23621 + }, + { + "epoch": 18.86741214057508, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23622 + }, + { + "epoch": 18.868210862619808, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 23623 + }, + { + "epoch": 18.869009584664536, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0745, + "step": 23624 + }, + { + "epoch": 18.869808306709267, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 23625 + }, + { + "epoch": 18.870607028753994, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 23626 + }, + { + "epoch": 18.871405750798722, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 23627 + }, + { + "epoch": 18.87220447284345, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 23628 + }, + { + "epoch": 18.873003194888177, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 23629 + }, + { + "epoch": 18.87380191693291, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 23630 + }, + { + "epoch": 18.874600638977636, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 23631 + }, + { + "epoch": 18.875399361022364, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 23632 + }, + { + "epoch": 18.87619808306709, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 23633 + }, + { + "epoch": 18.876996805111823, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 23634 + }, + { + "epoch": 18.87779552715655, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23635 + }, + { + "epoch": 18.878594249201278, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 23636 + }, + { + "epoch": 18.879392971246006, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 23637 + }, + { + "epoch": 18.880191693290733, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23638 + }, + { + "epoch": 18.880990415335464, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23639 + }, + { + "epoch": 18.881789137380192, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 23640 + }, + { + "epoch": 18.88258785942492, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23641 + }, + { + "epoch": 18.883386581469647, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 23642 + }, + { + "epoch": 18.88418530351438, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 23643 + }, + { + "epoch": 18.884984025559106, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 23644 + }, + { + "epoch": 18.885782747603834, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 23645 + }, + { + "epoch": 18.88658146964856, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 23646 + }, + { + "epoch": 18.88738019169329, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 23647 + }, + { + "epoch": 18.88817891373802, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 23648 + }, + { + "epoch": 18.888977635782748, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 23649 + }, + { + "epoch": 18.889776357827476, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 23650 + }, + { + "epoch": 18.890575079872203, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 23651 + }, + { + "epoch": 18.891373801916934, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 23652 + }, + { + "epoch": 18.892172523961662, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 23653 + }, + { + "epoch": 18.89297124600639, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 23654 + }, + { + "epoch": 18.893769968051117, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 23655 + }, + { + "epoch": 18.894568690095845, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 23656 + }, + { + "epoch": 18.895367412140576, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 23657 + }, + { + "epoch": 18.896166134185304, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 23658 + }, + { + "epoch": 18.89696485623003, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 23659 + }, + { + "epoch": 18.89776357827476, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 23660 + }, + { + "epoch": 18.89856230031949, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 23661 + }, + { + "epoch": 18.899361022364218, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 23662 + }, + { + "epoch": 18.900159744408946, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 23663 + }, + { + "epoch": 18.900958466453673, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 23664 + }, + { + "epoch": 18.9017571884984, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 23665 + }, + { + "epoch": 18.902555910543132, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23666 + }, + { + "epoch": 18.90335463258786, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 23667 + }, + { + "epoch": 18.904153354632587, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 23668 + }, + { + "epoch": 18.904952076677315, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 23669 + }, + { + "epoch": 18.905750798722046, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 23670 + }, + { + "epoch": 18.906549520766774, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 23671 + }, + { + "epoch": 18.9073482428115, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23672 + }, + { + "epoch": 18.90814696485623, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 23673 + }, + { + "epoch": 18.908945686900957, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 23674 + }, + { + "epoch": 18.909744408945688, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 23675 + }, + { + "epoch": 18.910543130990416, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 23676 + }, + { + "epoch": 18.911341853035143, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 23677 + }, + { + "epoch": 18.91214057507987, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 23678 + }, + { + "epoch": 18.912939297124602, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 23679 + }, + { + "epoch": 18.91373801916933, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23680 + }, + { + "epoch": 18.914536741214057, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 23681 + }, + { + "epoch": 18.915335463258785, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 23682 + }, + { + "epoch": 18.916134185303516, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 23683 + }, + { + "epoch": 18.916932907348244, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 23684 + }, + { + "epoch": 18.91773162939297, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 23685 + }, + { + "epoch": 18.9185303514377, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 23686 + }, + { + "epoch": 18.919329073482427, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 23687 + }, + { + "epoch": 18.920127795527158, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 23688 + }, + { + "epoch": 18.920926517571885, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 23689 + }, + { + "epoch": 18.921725239616613, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 23690 + }, + { + "epoch": 18.92252396166134, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 23691 + }, + { + "epoch": 18.923322683706072, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 23692 + }, + { + "epoch": 18.9241214057508, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23693 + }, + { + "epoch": 18.924920127795527, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 23694 + }, + { + "epoch": 18.925718849840255, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23695 + }, + { + "epoch": 18.926517571884983, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 23696 + }, + { + "epoch": 18.927316293929714, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 23697 + }, + { + "epoch": 18.92811501597444, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 23698 + }, + { + "epoch": 18.92891373801917, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 23699 + }, + { + "epoch": 18.929712460063897, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23700 + }, + { + "epoch": 18.930511182108628, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 23701 + }, + { + "epoch": 18.931309904153355, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 23702 + }, + { + "epoch": 18.932108626198083, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 23703 + }, + { + "epoch": 18.93290734824281, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 23704 + }, + { + "epoch": 18.93370607028754, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 23705 + }, + { + "epoch": 18.93450479233227, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 23706 + }, + { + "epoch": 18.935303514376997, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 23707 + }, + { + "epoch": 18.936102236421725, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 23708 + }, + { + "epoch": 18.936900958466452, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23709 + }, + { + "epoch": 18.937699680511184, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 23710 + }, + { + "epoch": 18.93849840255591, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 23711 + }, + { + "epoch": 18.93929712460064, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 23712 + }, + { + "epoch": 18.940095846645367, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 23713 + }, + { + "epoch": 18.940894568690094, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23714 + }, + { + "epoch": 18.941693290734825, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23715 + }, + { + "epoch": 18.942492012779553, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23716 + }, + { + "epoch": 18.94329073482428, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 23717 + }, + { + "epoch": 18.94408945686901, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 23718 + }, + { + "epoch": 18.94488817891374, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 23719 + }, + { + "epoch": 18.945686900958467, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23720 + }, + { + "epoch": 18.946485623003195, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 23721 + }, + { + "epoch": 18.947284345047922, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23722 + }, + { + "epoch": 18.94808306709265, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 23723 + }, + { + "epoch": 18.94888178913738, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 23724 + }, + { + "epoch": 18.94968051118211, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 23725 + }, + { + "epoch": 18.950479233226837, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 23726 + }, + { + "epoch": 18.951277955271564, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 23727 + }, + { + "epoch": 18.952076677316295, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 23728 + }, + { + "epoch": 18.952875399361023, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 23729 + }, + { + "epoch": 18.95367412140575, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 23730 + }, + { + "epoch": 18.95447284345048, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 23731 + }, + { + "epoch": 18.955271565495206, + "grad_norm": 0.318359375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 23732 + }, + { + "epoch": 18.956070287539937, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 23733 + }, + { + "epoch": 18.956869009584665, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 23734 + }, + { + "epoch": 18.957667731629392, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 23735 + }, + { + "epoch": 18.95846645367412, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 23736 + }, + { + "epoch": 18.95926517571885, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 23737 + }, + { + "epoch": 18.96006389776358, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 23738 + }, + { + "epoch": 18.960862619808307, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 23739 + }, + { + "epoch": 18.961661341853034, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 23740 + }, + { + "epoch": 18.962460063897765, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 23741 + }, + { + "epoch": 18.963258785942493, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 23742 + }, + { + "epoch": 18.96405750798722, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 23743 + }, + { + "epoch": 18.96485623003195, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23744 + }, + { + "epoch": 18.965654952076676, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 23745 + }, + { + "epoch": 18.966453674121407, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23746 + }, + { + "epoch": 18.967252396166135, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 23747 + }, + { + "epoch": 18.968051118210862, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 23748 + }, + { + "epoch": 18.96884984025559, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 23749 + }, + { + "epoch": 18.96964856230032, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 23750 + }, + { + "epoch": 18.97044728434505, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 23751 + }, + { + "epoch": 18.971246006389777, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 23752 + }, + { + "epoch": 18.972044728434504, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 23753 + }, + { + "epoch": 18.972843450479232, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 23754 + }, + { + "epoch": 18.973642172523963, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 23755 + }, + { + "epoch": 18.97444089456869, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 23756 + }, + { + "epoch": 18.97523961661342, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 23757 + }, + { + "epoch": 18.976038338658146, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 23758 + }, + { + "epoch": 18.976837060702877, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 23759 + }, + { + "epoch": 18.977635782747605, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 23760 + }, + { + "epoch": 18.978434504792332, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 23761 + }, + { + "epoch": 18.97923322683706, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 23762 + }, + { + "epoch": 18.980031948881788, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23763 + }, + { + "epoch": 18.98083067092652, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 23764 + }, + { + "epoch": 18.981629392971247, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23765 + }, + { + "epoch": 18.982428115015974, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 23766 + }, + { + "epoch": 18.983226837060702, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 23767 + }, + { + "epoch": 18.984025559105433, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 23768 + }, + { + "epoch": 18.98482428115016, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 23769 + }, + { + "epoch": 18.98562300319489, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 23770 + }, + { + "epoch": 18.986421725239616, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 23771 + }, + { + "epoch": 18.987220447284344, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 23772 + }, + { + "epoch": 18.988019169329075, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 23773 + }, + { + "epoch": 18.988817891373802, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 23774 + }, + { + "epoch": 18.98961661341853, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 23775 + }, + { + "epoch": 18.990415335463258, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23776 + }, + { + "epoch": 18.99121405750799, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23777 + }, + { + "epoch": 18.992012779552716, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 23778 + }, + { + "epoch": 18.992811501597444, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 23779 + }, + { + "epoch": 18.99361022364217, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 23780 + }, + { + "epoch": 18.9944089456869, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 23781 + }, + { + "epoch": 18.99520766773163, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 23782 + }, + { + "epoch": 18.99600638977636, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 23783 + }, + { + "epoch": 18.996805111821086, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 23784 + }, + { + "epoch": 18.997603833865814, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 23785 + }, + { + "epoch": 18.998402555910545, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 23786 + }, + { + "epoch": 18.999201277955272, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 23787 + }, + { + "epoch": 19.0, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 23788 + }, + { + "epoch": 19.000798722044728, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 23789 + }, + { + "epoch": 19.001597444089455, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23790 + }, + { + "epoch": 19.002396166134186, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 23791 + }, + { + "epoch": 19.003194888178914, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 23792 + }, + { + "epoch": 19.00399361022364, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 23793 + }, + { + "epoch": 19.00479233226837, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 23794 + }, + { + "epoch": 19.0055910543131, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 23795 + }, + { + "epoch": 19.00638977635783, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23796 + }, + { + "epoch": 19.007188498402556, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 23797 + }, + { + "epoch": 19.007987220447284, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 23798 + }, + { + "epoch": 19.00878594249201, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 23799 + }, + { + "epoch": 19.009584664536742, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 23800 + }, + { + "epoch": 19.01038338658147, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23801 + }, + { + "epoch": 19.011182108626198, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 23802 + }, + { + "epoch": 19.011980830670925, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 23803 + }, + { + "epoch": 19.012779552715656, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 23804 + }, + { + "epoch": 19.013578274760384, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 23805 + }, + { + "epoch": 19.01437699680511, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0757, + "step": 23806 + }, + { + "epoch": 19.01517571884984, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 23807 + }, + { + "epoch": 19.015974440894567, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 23808 + }, + { + "epoch": 19.016773162939298, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 23809 + }, + { + "epoch": 19.017571884984026, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 23810 + }, + { + "epoch": 19.018370607028753, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 23811 + }, + { + "epoch": 19.01916932907348, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23812 + }, + { + "epoch": 19.019968051118212, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 23813 + }, + { + "epoch": 19.02076677316294, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 23814 + }, + { + "epoch": 19.021565495207668, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 23815 + }, + { + "epoch": 19.022364217252395, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 23816 + }, + { + "epoch": 19.023162939297123, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 23817 + }, + { + "epoch": 19.023961661341854, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 23818 + }, + { + "epoch": 19.02476038338658, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 23819 + }, + { + "epoch": 19.02555910543131, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 23820 + }, + { + "epoch": 19.026357827476037, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 23821 + }, + { + "epoch": 19.027156549520768, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23822 + }, + { + "epoch": 19.027955271565496, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 23823 + }, + { + "epoch": 19.028753993610223, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 23824 + }, + { + "epoch": 19.02955271565495, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 23825 + }, + { + "epoch": 19.03035143769968, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 23826 + }, + { + "epoch": 19.03115015974441, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 23827 + }, + { + "epoch": 19.031948881789138, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 23828 + }, + { + "epoch": 19.032747603833865, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23829 + }, + { + "epoch": 19.033546325878593, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 23830 + }, + { + "epoch": 19.034345047923324, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 23831 + }, + { + "epoch": 19.03514376996805, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 23832 + }, + { + "epoch": 19.03594249201278, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 23833 + }, + { + "epoch": 19.036741214057507, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 23834 + }, + { + "epoch": 19.037539936102238, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 23835 + }, + { + "epoch": 19.038338658146966, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 23836 + }, + { + "epoch": 19.039137380191693, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 23837 + }, + { + "epoch": 19.03993610223642, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 23838 + }, + { + "epoch": 19.04073482428115, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 23839 + }, + { + "epoch": 19.04153354632588, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 23840 + }, + { + "epoch": 19.042332268370608, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 23841 + }, + { + "epoch": 19.043130990415335, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 23842 + }, + { + "epoch": 19.043929712460063, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 23843 + }, + { + "epoch": 19.044728434504794, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 23844 + }, + { + "epoch": 19.04552715654952, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 23845 + }, + { + "epoch": 19.04632587859425, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 23846 + }, + { + "epoch": 19.047124600638977, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 23847 + }, + { + "epoch": 19.047923322683705, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 23848 + }, + { + "epoch": 19.048722044728436, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 23849 + }, + { + "epoch": 19.049520766773163, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 23850 + }, + { + "epoch": 19.05031948881789, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 23851 + }, + { + "epoch": 19.05111821086262, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 23852 + }, + { + "epoch": 19.05191693290735, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 23853 + }, + { + "epoch": 19.052715654952078, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 23854 + }, + { + "epoch": 19.053514376996805, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 23855 + }, + { + "epoch": 19.054313099041533, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 23856 + }, + { + "epoch": 19.05511182108626, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 23857 + }, + { + "epoch": 19.05591054313099, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 23858 + }, + { + "epoch": 19.05670926517572, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23859 + }, + { + "epoch": 19.057507987220447, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 23860 + }, + { + "epoch": 19.058306709265175, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 23861 + }, + { + "epoch": 19.059105431309906, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 23862 + }, + { + "epoch": 19.059904153354633, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 23863 + }, + { + "epoch": 19.06070287539936, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 23864 + }, + { + "epoch": 19.06150159744409, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 23865 + }, + { + "epoch": 19.062300319488816, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 23866 + }, + { + "epoch": 19.063099041533548, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 23867 + }, + { + "epoch": 19.063897763578275, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 23868 + }, + { + "epoch": 19.064696485623003, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 23869 + }, + { + "epoch": 19.06549520766773, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 23870 + }, + { + "epoch": 19.06629392971246, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 23871 + }, + { + "epoch": 19.06709265175719, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 23872 + }, + { + "epoch": 19.067891373801917, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 23873 + }, + { + "epoch": 19.068690095846645, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 23874 + }, + { + "epoch": 19.069488817891372, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 23875 + }, + { + "epoch": 19.070287539936103, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 23876 + }, + { + "epoch": 19.07108626198083, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 23877 + }, + { + "epoch": 19.07188498402556, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23878 + }, + { + "epoch": 19.072683706070286, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 23879 + }, + { + "epoch": 19.073482428115017, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 23880 + }, + { + "epoch": 19.074281150159745, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 23881 + }, + { + "epoch": 19.075079872204473, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 23882 + }, + { + "epoch": 19.0758785942492, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 23883 + }, + { + "epoch": 19.076677316293928, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 23884 + }, + { + "epoch": 19.07747603833866, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 23885 + }, + { + "epoch": 19.078274760383387, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 23886 + }, + { + "epoch": 19.079073482428115, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 23887 + }, + { + "epoch": 19.079872204472842, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 23888 + }, + { + "epoch": 19.080670926517573, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 23889 + }, + { + "epoch": 19.0814696485623, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 23890 + }, + { + "epoch": 19.08226837060703, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 23891 + }, + { + "epoch": 19.083067092651756, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23892 + }, + { + "epoch": 19.083865814696484, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 23893 + }, + { + "epoch": 19.084664536741215, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 23894 + }, + { + "epoch": 19.085463258785943, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 23895 + }, + { + "epoch": 19.08626198083067, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 23896 + }, + { + "epoch": 19.087060702875398, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 23897 + }, + { + "epoch": 19.08785942492013, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 23898 + }, + { + "epoch": 19.088658146964857, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 23899 + }, + { + "epoch": 19.089456869009584, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 23900 + }, + { + "epoch": 19.090255591054312, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 23901 + }, + { + "epoch": 19.09105431309904, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 23902 + }, + { + "epoch": 19.09185303514377, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 23903 + }, + { + "epoch": 19.0926517571885, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 23904 + }, + { + "epoch": 19.093450479233226, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 23905 + }, + { + "epoch": 19.094249201277954, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 23906 + }, + { + "epoch": 19.095047923322685, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 23907 + }, + { + "epoch": 19.095846645367413, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 23908 + }, + { + "epoch": 19.09664536741214, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 23909 + }, + { + "epoch": 19.097444089456868, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 23910 + }, + { + "epoch": 19.0982428115016, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 23911 + }, + { + "epoch": 19.099041533546327, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 23912 + }, + { + "epoch": 19.099840255591054, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 23913 + }, + { + "epoch": 19.100638977635782, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23914 + }, + { + "epoch": 19.10143769968051, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 23915 + }, + { + "epoch": 19.10223642172524, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 23916 + }, + { + "epoch": 19.10303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 23917 + }, + { + "epoch": 19.103833865814696, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 23918 + }, + { + "epoch": 19.104632587859424, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 23919 + }, + { + "epoch": 19.105431309904155, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 23920 + }, + { + "epoch": 19.106230031948883, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 23921 + }, + { + "epoch": 19.10702875399361, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 23922 + }, + { + "epoch": 19.107827476038338, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 23923 + }, + { + "epoch": 19.108626198083066, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 23924 + }, + { + "epoch": 19.109424920127797, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 23925 + }, + { + "epoch": 19.110223642172524, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 23926 + }, + { + "epoch": 19.111022364217252, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 23927 + }, + { + "epoch": 19.11182108626198, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 23928 + }, + { + "epoch": 19.11261980830671, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 23929 + }, + { + "epoch": 19.11341853035144, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 23930 + }, + { + "epoch": 19.114217252396166, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 23931 + }, + { + "epoch": 19.115015974440894, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 23932 + }, + { + "epoch": 19.11581469648562, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23933 + }, + { + "epoch": 19.116613418530353, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 23934 + }, + { + "epoch": 19.11741214057508, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 23935 + }, + { + "epoch": 19.118210862619808, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 23936 + }, + { + "epoch": 19.119009584664536, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 23937 + }, + { + "epoch": 19.119808306709267, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 23938 + }, + { + "epoch": 19.120607028753994, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 23939 + }, + { + "epoch": 19.121405750798722, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23940 + }, + { + "epoch": 19.12220447284345, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 23941 + }, + { + "epoch": 19.123003194888177, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 23942 + }, + { + "epoch": 19.12380191693291, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 23943 + }, + { + "epoch": 19.124600638977636, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 23944 + }, + { + "epoch": 19.125399361022364, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 23945 + }, + { + "epoch": 19.12619808306709, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 23946 + }, + { + "epoch": 19.126996805111823, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 23947 + }, + { + "epoch": 19.12779552715655, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 23948 + }, + { + "epoch": 19.128594249201278, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 23949 + }, + { + "epoch": 19.129392971246006, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 23950 + }, + { + "epoch": 19.130191693290733, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 23951 + }, + { + "epoch": 19.130990415335464, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 23952 + }, + { + "epoch": 19.131789137380192, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 23953 + }, + { + "epoch": 19.13258785942492, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 23954 + }, + { + "epoch": 19.133386581469647, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 23955 + }, + { + "epoch": 19.13418530351438, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 23956 + }, + { + "epoch": 19.134984025559106, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 23957 + }, + { + "epoch": 19.135782747603834, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 23958 + }, + { + "epoch": 19.13658146964856, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 23959 + }, + { + "epoch": 19.13738019169329, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 23960 + }, + { + "epoch": 19.13817891373802, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 23961 + }, + { + "epoch": 19.138977635782748, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 23962 + }, + { + "epoch": 19.139776357827476, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 23963 + }, + { + "epoch": 19.140575079872203, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 23964 + }, + { + "epoch": 19.141373801916934, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 23965 + }, + { + "epoch": 19.142172523961662, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 23966 + }, + { + "epoch": 19.14297124600639, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 23967 + }, + { + "epoch": 19.143769968051117, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 23968 + }, + { + "epoch": 19.144568690095845, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 23969 + }, + { + "epoch": 19.145367412140576, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 23970 + }, + { + "epoch": 19.146166134185304, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 23971 + }, + { + "epoch": 19.14696485623003, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 23972 + }, + { + "epoch": 19.14776357827476, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 23973 + }, + { + "epoch": 19.14856230031949, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 23974 + }, + { + "epoch": 19.149361022364218, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 23975 + }, + { + "epoch": 19.150159744408946, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 23976 + }, + { + "epoch": 19.150958466453673, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23977 + }, + { + "epoch": 19.1517571884984, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23978 + }, + { + "epoch": 19.152555910543132, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 23979 + }, + { + "epoch": 19.15335463258786, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 23980 + }, + { + "epoch": 19.154153354632587, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 23981 + }, + { + "epoch": 19.154952076677315, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 23982 + }, + { + "epoch": 19.155750798722046, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 23983 + }, + { + "epoch": 19.156549520766774, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 23984 + }, + { + "epoch": 19.1573482428115, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 23985 + }, + { + "epoch": 19.15814696485623, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 23986 + }, + { + "epoch": 19.15894568690096, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 23987 + }, + { + "epoch": 19.159744408945688, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 23988 + }, + { + "epoch": 19.160543130990416, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 23989 + }, + { + "epoch": 19.161341853035143, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 23990 + }, + { + "epoch": 19.16214057507987, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 23991 + }, + { + "epoch": 19.162939297124602, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 23992 + }, + { + "epoch": 19.16373801916933, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 23993 + }, + { + "epoch": 19.164536741214057, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 23994 + }, + { + "epoch": 19.165335463258785, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 23995 + }, + { + "epoch": 19.166134185303516, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 23996 + }, + { + "epoch": 19.166932907348244, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 23997 + }, + { + "epoch": 19.16773162939297, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 23998 + }, + { + "epoch": 19.1685303514377, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 23999 + }, + { + "epoch": 19.169329073482427, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 24000 + }, + { + "epoch": 19.170127795527158, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 24001 + }, + { + "epoch": 19.170926517571885, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 24002 + }, + { + "epoch": 19.171725239616613, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 24003 + }, + { + "epoch": 19.17252396166134, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 24004 + }, + { + "epoch": 19.173322683706072, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 24005 + }, + { + "epoch": 19.1741214057508, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 24006 + }, + { + "epoch": 19.174920127795527, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24007 + }, + { + "epoch": 19.175718849840255, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 24008 + }, + { + "epoch": 19.176517571884983, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 24009 + }, + { + "epoch": 19.177316293929714, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 24010 + }, + { + "epoch": 19.17811501597444, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 24011 + }, + { + "epoch": 19.17891373801917, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 24012 + }, + { + "epoch": 19.179712460063897, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 24013 + }, + { + "epoch": 19.180511182108628, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 24014 + }, + { + "epoch": 19.181309904153355, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 24015 + }, + { + "epoch": 19.182108626198083, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 24016 + }, + { + "epoch": 19.18290734824281, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 24017 + }, + { + "epoch": 19.18370607028754, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 24018 + }, + { + "epoch": 19.18450479233227, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 24019 + }, + { + "epoch": 19.185303514376997, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 24020 + }, + { + "epoch": 19.186102236421725, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 24021 + }, + { + "epoch": 19.186900958466452, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 24022 + }, + { + "epoch": 19.187699680511184, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24023 + }, + { + "epoch": 19.18849840255591, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 24024 + }, + { + "epoch": 19.18929712460064, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 24025 + }, + { + "epoch": 19.190095846645367, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 24026 + }, + { + "epoch": 19.190894568690094, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 24027 + }, + { + "epoch": 19.191693290734825, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 24028 + }, + { + "epoch": 19.192492012779553, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 24029 + }, + { + "epoch": 19.19329073482428, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 24030 + }, + { + "epoch": 19.19408945686901, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 24031 + }, + { + "epoch": 19.19488817891374, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 24032 + }, + { + "epoch": 19.195686900958467, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 24033 + }, + { + "epoch": 19.196485623003195, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 24034 + }, + { + "epoch": 19.197284345047922, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 24035 + }, + { + "epoch": 19.19808306709265, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 24036 + }, + { + "epoch": 19.19888178913738, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 24037 + }, + { + "epoch": 19.19968051118211, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 24038 + }, + { + "epoch": 19.200479233226837, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 24039 + }, + { + "epoch": 19.201277955271564, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 24040 + }, + { + "epoch": 19.202076677316295, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 24041 + }, + { + "epoch": 19.202875399361023, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 24042 + }, + { + "epoch": 19.20367412140575, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 24043 + }, + { + "epoch": 19.20447284345048, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 24044 + }, + { + "epoch": 19.205271565495206, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24045 + }, + { + "epoch": 19.206070287539937, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24046 + }, + { + "epoch": 19.206869009584665, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 24047 + }, + { + "epoch": 19.207667731629392, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 24048 + }, + { + "epoch": 19.20846645367412, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 24049 + }, + { + "epoch": 19.20926517571885, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 24050 + }, + { + "epoch": 19.21006389776358, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 24051 + }, + { + "epoch": 19.210862619808307, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 24052 + }, + { + "epoch": 19.211661341853034, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 24053 + }, + { + "epoch": 19.212460063897762, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 24054 + }, + { + "epoch": 19.213258785942493, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 24055 + }, + { + "epoch": 19.21405750798722, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 24056 + }, + { + "epoch": 19.21485623003195, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 24057 + }, + { + "epoch": 19.215654952076676, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 24058 + }, + { + "epoch": 19.216453674121407, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 24059 + }, + { + "epoch": 19.217252396166135, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 24060 + }, + { + "epoch": 19.218051118210862, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 24061 + }, + { + "epoch": 19.21884984025559, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 24062 + }, + { + "epoch": 19.21964856230032, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24063 + }, + { + "epoch": 19.22044728434505, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 24064 + }, + { + "epoch": 19.221246006389777, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 24065 + }, + { + "epoch": 19.222044728434504, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 24066 + }, + { + "epoch": 19.222843450479232, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24067 + }, + { + "epoch": 19.223642172523963, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 24068 + }, + { + "epoch": 19.22444089456869, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 24069 + }, + { + "epoch": 19.22523961661342, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 24070 + }, + { + "epoch": 19.226038338658146, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 24071 + }, + { + "epoch": 19.226837060702877, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 24072 + }, + { + "epoch": 19.227635782747605, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 24073 + }, + { + "epoch": 19.228434504792332, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 24074 + }, + { + "epoch": 19.22923322683706, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 24075 + }, + { + "epoch": 19.230031948881788, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 24076 + }, + { + "epoch": 19.23083067092652, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 24077 + }, + { + "epoch": 19.231629392971247, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 24078 + }, + { + "epoch": 19.232428115015974, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 24079 + }, + { + "epoch": 19.233226837060702, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 24080 + }, + { + "epoch": 19.234025559105433, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 24081 + }, + { + "epoch": 19.23482428115016, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 24082 + }, + { + "epoch": 19.23562300319489, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 24083 + }, + { + "epoch": 19.236421725239616, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 24084 + }, + { + "epoch": 19.237220447284344, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 24085 + }, + { + "epoch": 19.238019169329075, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 24086 + }, + { + "epoch": 19.238817891373802, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 24087 + }, + { + "epoch": 19.23961661341853, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24088 + }, + { + "epoch": 19.240415335463258, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 24089 + }, + { + "epoch": 19.24121405750799, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24090 + }, + { + "epoch": 19.242012779552716, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 24091 + }, + { + "epoch": 19.242811501597444, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 24092 + }, + { + "epoch": 19.24361022364217, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 24093 + }, + { + "epoch": 19.2444089456869, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 24094 + }, + { + "epoch": 19.24520766773163, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 24095 + }, + { + "epoch": 19.24600638977636, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 24096 + }, + { + "epoch": 19.246805111821086, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 24097 + }, + { + "epoch": 19.247603833865814, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 24098 + }, + { + "epoch": 19.248402555910545, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 24099 + }, + { + "epoch": 19.249201277955272, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24100 + }, + { + "epoch": 19.25, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24101 + }, + { + "epoch": 19.250798722044728, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24102 + }, + { + "epoch": 19.251597444089455, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24103 + }, + { + "epoch": 19.252396166134186, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 24104 + }, + { + "epoch": 19.253194888178914, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 24105 + }, + { + "epoch": 19.25399361022364, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 24106 + }, + { + "epoch": 19.25479233226837, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 24107 + }, + { + "epoch": 19.2555910543131, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 24108 + }, + { + "epoch": 19.25638977635783, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 24109 + }, + { + "epoch": 19.257188498402556, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 24110 + }, + { + "epoch": 19.257987220447284, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 24111 + }, + { + "epoch": 19.25878594249201, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 24112 + }, + { + "epoch": 19.259584664536742, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 24113 + }, + { + "epoch": 19.26038338658147, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 24114 + }, + { + "epoch": 19.261182108626198, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 24115 + }, + { + "epoch": 19.261980830670925, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 24116 + }, + { + "epoch": 19.262779552715656, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 24117 + }, + { + "epoch": 19.263578274760384, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 24118 + }, + { + "epoch": 19.26437699680511, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 24119 + }, + { + "epoch": 19.26517571884984, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 24120 + }, + { + "epoch": 19.265974440894567, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 24121 + }, + { + "epoch": 19.266773162939298, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 24122 + }, + { + "epoch": 19.267571884984026, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 24123 + }, + { + "epoch": 19.268370607028753, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 24124 + }, + { + "epoch": 19.26916932907348, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24125 + }, + { + "epoch": 19.269968051118212, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 24126 + }, + { + "epoch": 19.27076677316294, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 24127 + }, + { + "epoch": 19.271565495207668, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 24128 + }, + { + "epoch": 19.272364217252395, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 24129 + }, + { + "epoch": 19.273162939297123, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 24130 + }, + { + "epoch": 19.273961661341854, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 24131 + }, + { + "epoch": 19.27476038338658, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 24132 + }, + { + "epoch": 19.27555910543131, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 24133 + }, + { + "epoch": 19.276357827476037, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 24134 + }, + { + "epoch": 19.277156549520768, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 24135 + }, + { + "epoch": 19.277955271565496, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24136 + }, + { + "epoch": 19.278753993610223, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 24137 + }, + { + "epoch": 19.27955271565495, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 24138 + }, + { + "epoch": 19.28035143769968, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 24139 + }, + { + "epoch": 19.28115015974441, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 24140 + }, + { + "epoch": 19.281948881789138, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 24141 + }, + { + "epoch": 19.282747603833865, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 24142 + }, + { + "epoch": 19.283546325878593, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 24143 + }, + { + "epoch": 19.284345047923324, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 24144 + }, + { + "epoch": 19.28514376996805, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 24145 + }, + { + "epoch": 19.28594249201278, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 24146 + }, + { + "epoch": 19.286741214057507, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 24147 + }, + { + "epoch": 19.287539936102238, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 24148 + }, + { + "epoch": 19.288338658146966, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24149 + }, + { + "epoch": 19.289137380191693, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 24150 + }, + { + "epoch": 19.28993610223642, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 24151 + }, + { + "epoch": 19.29073482428115, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 24152 + }, + { + "epoch": 19.29153354632588, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 24153 + }, + { + "epoch": 19.292332268370608, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 24154 + }, + { + "epoch": 19.293130990415335, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 24155 + }, + { + "epoch": 19.293929712460063, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 24156 + }, + { + "epoch": 19.294728434504794, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 24157 + }, + { + "epoch": 19.29552715654952, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 24158 + }, + { + "epoch": 19.29632587859425, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 24159 + }, + { + "epoch": 19.297124600638977, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 24160 + }, + { + "epoch": 19.297923322683705, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 24161 + }, + { + "epoch": 19.298722044728436, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 24162 + }, + { + "epoch": 19.299520766773163, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 24163 + }, + { + "epoch": 19.30031948881789, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 24164 + }, + { + "epoch": 19.30111821086262, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 24165 + }, + { + "epoch": 19.30191693290735, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24166 + }, + { + "epoch": 19.302715654952078, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 24167 + }, + { + "epoch": 19.303514376996805, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 24168 + }, + { + "epoch": 19.304313099041533, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 24169 + }, + { + "epoch": 19.30511182108626, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 24170 + }, + { + "epoch": 19.30591054313099, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 24171 + }, + { + "epoch": 19.30670926517572, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24172 + }, + { + "epoch": 19.307507987220447, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 24173 + }, + { + "epoch": 19.308306709265175, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 24174 + }, + { + "epoch": 19.309105431309906, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 24175 + }, + { + "epoch": 19.309904153354633, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 24176 + }, + { + "epoch": 19.31070287539936, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 24177 + }, + { + "epoch": 19.31150159744409, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24178 + }, + { + "epoch": 19.312300319488816, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 24179 + }, + { + "epoch": 19.313099041533548, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 24180 + }, + { + "epoch": 19.313897763578275, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 24181 + }, + { + "epoch": 19.314696485623003, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 24182 + }, + { + "epoch": 19.31549520766773, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24183 + }, + { + "epoch": 19.31629392971246, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 24184 + }, + { + "epoch": 19.31709265175719, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 24185 + }, + { + "epoch": 19.317891373801917, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 24186 + }, + { + "epoch": 19.318690095846645, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 24187 + }, + { + "epoch": 19.319488817891372, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 24188 + }, + { + "epoch": 19.320287539936103, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 24189 + }, + { + "epoch": 19.32108626198083, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 24190 + }, + { + "epoch": 19.32188498402556, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 24191 + }, + { + "epoch": 19.322683706070286, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24192 + }, + { + "epoch": 19.323482428115017, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 24193 + }, + { + "epoch": 19.324281150159745, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 24194 + }, + { + "epoch": 19.325079872204473, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 24195 + }, + { + "epoch": 19.3258785942492, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 24196 + }, + { + "epoch": 19.326677316293928, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 24197 + }, + { + "epoch": 19.32747603833866, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 24198 + }, + { + "epoch": 19.328274760383387, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 24199 + }, + { + "epoch": 19.329073482428115, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 24200 + }, + { + "epoch": 19.329872204472842, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 24201 + }, + { + "epoch": 19.330670926517573, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 24202 + }, + { + "epoch": 19.3314696485623, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24203 + }, + { + "epoch": 19.33226837060703, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 24204 + }, + { + "epoch": 19.333067092651756, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 24205 + }, + { + "epoch": 19.333865814696484, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 24206 + }, + { + "epoch": 19.334664536741215, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 24207 + }, + { + "epoch": 19.335463258785943, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 24208 + }, + { + "epoch": 19.33626198083067, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24209 + }, + { + "epoch": 19.337060702875398, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 24210 + }, + { + "epoch": 19.33785942492013, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24211 + }, + { + "epoch": 19.338658146964857, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24212 + }, + { + "epoch": 19.339456869009584, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 24213 + }, + { + "epoch": 19.340255591054312, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24214 + }, + { + "epoch": 19.341054313099043, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 24215 + }, + { + "epoch": 19.34185303514377, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 24216 + }, + { + "epoch": 19.3426517571885, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 24217 + }, + { + "epoch": 19.343450479233226, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 24218 + }, + { + "epoch": 19.344249201277954, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 24219 + }, + { + "epoch": 19.345047923322685, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 24220 + }, + { + "epoch": 19.345846645367413, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 24221 + }, + { + "epoch": 19.34664536741214, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 24222 + }, + { + "epoch": 19.347444089456868, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 24223 + }, + { + "epoch": 19.3482428115016, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 24224 + }, + { + "epoch": 19.349041533546327, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 24225 + }, + { + "epoch": 19.349840255591054, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 24226 + }, + { + "epoch": 19.350638977635782, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 24227 + }, + { + "epoch": 19.35143769968051, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 24228 + }, + { + "epoch": 19.35223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 24229 + }, + { + "epoch": 19.35303514376997, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 24230 + }, + { + "epoch": 19.353833865814696, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 24231 + }, + { + "epoch": 19.354632587859424, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 24232 + }, + { + "epoch": 19.355431309904155, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 24233 + }, + { + "epoch": 19.356230031948883, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 24234 + }, + { + "epoch": 19.35702875399361, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 24235 + }, + { + "epoch": 19.357827476038338, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 24236 + }, + { + "epoch": 19.358626198083066, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 24237 + }, + { + "epoch": 19.359424920127797, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 24238 + }, + { + "epoch": 19.360223642172524, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 24239 + }, + { + "epoch": 19.361022364217252, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24240 + }, + { + "epoch": 19.36182108626198, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 24241 + }, + { + "epoch": 19.36261980830671, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 24242 + }, + { + "epoch": 19.36341853035144, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 24243 + }, + { + "epoch": 19.364217252396166, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 24244 + }, + { + "epoch": 19.365015974440894, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 24245 + }, + { + "epoch": 19.36581469648562, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 24246 + }, + { + "epoch": 19.366613418530353, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24247 + }, + { + "epoch": 19.36741214057508, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 24248 + }, + { + "epoch": 19.368210862619808, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 24249 + }, + { + "epoch": 19.369009584664536, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 24250 + }, + { + "epoch": 19.369808306709267, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 24251 + }, + { + "epoch": 19.370607028753994, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 24252 + }, + { + "epoch": 19.371405750798722, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 24253 + }, + { + "epoch": 19.37220447284345, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 24254 + }, + { + "epoch": 19.373003194888177, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1105, + "step": 24255 + }, + { + "epoch": 19.37380191693291, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 24256 + }, + { + "epoch": 19.374600638977636, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 24257 + }, + { + "epoch": 19.375399361022364, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 24258 + }, + { + "epoch": 19.37619808306709, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 24259 + }, + { + "epoch": 19.376996805111823, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 24260 + }, + { + "epoch": 19.37779552715655, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 24261 + }, + { + "epoch": 19.378594249201278, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 24262 + }, + { + "epoch": 19.379392971246006, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 24263 + }, + { + "epoch": 19.380191693290733, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 24264 + }, + { + "epoch": 19.380990415335464, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 24265 + }, + { + "epoch": 19.381789137380192, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24266 + }, + { + "epoch": 19.38258785942492, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 24267 + }, + { + "epoch": 19.383386581469647, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 24268 + }, + { + "epoch": 19.38418530351438, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 24269 + }, + { + "epoch": 19.384984025559106, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 24270 + }, + { + "epoch": 19.385782747603834, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 24271 + }, + { + "epoch": 19.38658146964856, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 24272 + }, + { + "epoch": 19.38738019169329, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 24273 + }, + { + "epoch": 19.38817891373802, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 24274 + }, + { + "epoch": 19.388977635782748, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 24275 + }, + { + "epoch": 19.389776357827476, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 24276 + }, + { + "epoch": 19.390575079872203, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 24277 + }, + { + "epoch": 19.391373801916934, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 24278 + }, + { + "epoch": 19.392172523961662, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 24279 + }, + { + "epoch": 19.39297124600639, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 24280 + }, + { + "epoch": 19.393769968051117, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 24281 + }, + { + "epoch": 19.394568690095845, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 24282 + }, + { + "epoch": 19.395367412140576, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24283 + }, + { + "epoch": 19.396166134185304, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 24284 + }, + { + "epoch": 19.39696485623003, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 24285 + }, + { + "epoch": 19.39776357827476, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 24286 + }, + { + "epoch": 19.39856230031949, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 24287 + }, + { + "epoch": 19.399361022364218, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24288 + }, + { + "epoch": 19.400159744408946, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 24289 + }, + { + "epoch": 19.400958466453673, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 24290 + }, + { + "epoch": 19.4017571884984, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 24291 + }, + { + "epoch": 19.402555910543132, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 24292 + }, + { + "epoch": 19.40335463258786, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 24293 + }, + { + "epoch": 19.404153354632587, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 24294 + }, + { + "epoch": 19.404952076677315, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24295 + }, + { + "epoch": 19.405750798722046, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 24296 + }, + { + "epoch": 19.406549520766774, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 24297 + }, + { + "epoch": 19.4073482428115, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 24298 + }, + { + "epoch": 19.40814696485623, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 24299 + }, + { + "epoch": 19.408945686900957, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 24300 + }, + { + "epoch": 19.409744408945688, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 24301 + }, + { + "epoch": 19.410543130990416, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 24302 + }, + { + "epoch": 19.411341853035143, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24303 + }, + { + "epoch": 19.41214057507987, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 24304 + }, + { + "epoch": 19.412939297124602, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 24305 + }, + { + "epoch": 19.41373801916933, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 24306 + }, + { + "epoch": 19.414536741214057, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 24307 + }, + { + "epoch": 19.415335463258785, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 24308 + }, + { + "epoch": 19.416134185303516, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 24309 + }, + { + "epoch": 19.416932907348244, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 24310 + }, + { + "epoch": 19.41773162939297, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 24311 + }, + { + "epoch": 19.4185303514377, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 24312 + }, + { + "epoch": 19.419329073482427, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24313 + }, + { + "epoch": 19.420127795527158, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 24314 + }, + { + "epoch": 19.420926517571885, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 24315 + }, + { + "epoch": 19.421725239616613, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 24316 + }, + { + "epoch": 19.42252396166134, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 24317 + }, + { + "epoch": 19.423322683706072, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 24318 + }, + { + "epoch": 19.4241214057508, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 24319 + }, + { + "epoch": 19.424920127795527, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 24320 + }, + { + "epoch": 19.425718849840255, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 24321 + }, + { + "epoch": 19.426517571884983, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 24322 + }, + { + "epoch": 19.427316293929714, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 24323 + }, + { + "epoch": 19.42811501597444, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 24324 + }, + { + "epoch": 19.42891373801917, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 24325 + }, + { + "epoch": 19.429712460063897, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 24326 + }, + { + "epoch": 19.430511182108628, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 24327 + }, + { + "epoch": 19.431309904153355, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 24328 + }, + { + "epoch": 19.432108626198083, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 24329 + }, + { + "epoch": 19.43290734824281, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 24330 + }, + { + "epoch": 19.43370607028754, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 24331 + }, + { + "epoch": 19.43450479233227, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 24332 + }, + { + "epoch": 19.435303514376997, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 24333 + }, + { + "epoch": 19.436102236421725, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 24334 + }, + { + "epoch": 19.436900958466452, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 24335 + }, + { + "epoch": 19.437699680511184, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 24336 + }, + { + "epoch": 19.43849840255591, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 24337 + }, + { + "epoch": 19.43929712460064, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24338 + }, + { + "epoch": 19.440095846645367, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 24339 + }, + { + "epoch": 19.440894568690094, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24340 + }, + { + "epoch": 19.441693290734825, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 24341 + }, + { + "epoch": 19.442492012779553, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24342 + }, + { + "epoch": 19.44329073482428, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 24343 + }, + { + "epoch": 19.44408945686901, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 24344 + }, + { + "epoch": 19.44488817891374, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 24345 + }, + { + "epoch": 19.445686900958467, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 24346 + }, + { + "epoch": 19.446485623003195, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 24347 + }, + { + "epoch": 19.447284345047922, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 24348 + }, + { + "epoch": 19.44808306709265, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 24349 + }, + { + "epoch": 19.44888178913738, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 24350 + }, + { + "epoch": 19.44968051118211, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24351 + }, + { + "epoch": 19.450479233226837, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 24352 + }, + { + "epoch": 19.451277955271564, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24353 + }, + { + "epoch": 19.452076677316295, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 24354 + }, + { + "epoch": 19.452875399361023, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 24355 + }, + { + "epoch": 19.45367412140575, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 24356 + }, + { + "epoch": 19.45447284345048, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 24357 + }, + { + "epoch": 19.455271565495206, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 24358 + }, + { + "epoch": 19.456070287539937, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 24359 + }, + { + "epoch": 19.456869009584665, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 24360 + }, + { + "epoch": 19.457667731629392, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 24361 + }, + { + "epoch": 19.45846645367412, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 24362 + }, + { + "epoch": 19.45926517571885, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 24363 + }, + { + "epoch": 19.46006389776358, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 24364 + }, + { + "epoch": 19.460862619808307, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 24365 + }, + { + "epoch": 19.461661341853034, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 24366 + }, + { + "epoch": 19.462460063897762, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 24367 + }, + { + "epoch": 19.463258785942493, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 24368 + }, + { + "epoch": 19.46405750798722, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 24369 + }, + { + "epoch": 19.46485623003195, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 24370 + }, + { + "epoch": 19.465654952076676, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 24371 + }, + { + "epoch": 19.466453674121407, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 24372 + }, + { + "epoch": 19.467252396166135, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 24373 + }, + { + "epoch": 19.468051118210862, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 24374 + }, + { + "epoch": 19.46884984025559, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 24375 + }, + { + "epoch": 19.46964856230032, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 24376 + }, + { + "epoch": 19.47044728434505, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 24377 + }, + { + "epoch": 19.471246006389777, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 24378 + }, + { + "epoch": 19.472044728434504, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 24379 + }, + { + "epoch": 19.472843450479232, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 24380 + }, + { + "epoch": 19.473642172523963, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 24381 + }, + { + "epoch": 19.47444089456869, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 24382 + }, + { + "epoch": 19.47523961661342, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 24383 + }, + { + "epoch": 19.476038338658146, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 24384 + }, + { + "epoch": 19.476837060702877, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 24385 + }, + { + "epoch": 19.477635782747605, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 24386 + }, + { + "epoch": 19.478434504792332, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 24387 + }, + { + "epoch": 19.47923322683706, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 24388 + }, + { + "epoch": 19.480031948881788, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 24389 + }, + { + "epoch": 19.48083067092652, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 24390 + }, + { + "epoch": 19.481629392971247, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 24391 + }, + { + "epoch": 19.482428115015974, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 24392 + }, + { + "epoch": 19.483226837060702, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 24393 + }, + { + "epoch": 19.484025559105433, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 24394 + }, + { + "epoch": 19.48482428115016, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 24395 + }, + { + "epoch": 19.48562300319489, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 24396 + }, + { + "epoch": 19.486421725239616, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 24397 + }, + { + "epoch": 19.487220447284344, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 24398 + }, + { + "epoch": 19.488019169329075, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 24399 + }, + { + "epoch": 19.488817891373802, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 24400 + }, + { + "epoch": 19.48961661341853, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 24401 + }, + { + "epoch": 19.490415335463258, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 24402 + }, + { + "epoch": 19.49121405750799, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0756, + "step": 24403 + }, + { + "epoch": 19.492012779552716, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 24404 + }, + { + "epoch": 19.492811501597444, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 24405 + }, + { + "epoch": 19.49361022364217, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 24406 + }, + { + "epoch": 19.4944089456869, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 24407 + }, + { + "epoch": 19.49520766773163, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 24408 + }, + { + "epoch": 19.49600638977636, + "grad_norm": 0.09033203125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 24409 + }, + { + "epoch": 19.496805111821086, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 24410 + }, + { + "epoch": 19.497603833865814, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 24411 + }, + { + "epoch": 19.498402555910545, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 24412 + }, + { + "epoch": 19.499201277955272, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 24413 + }, + { + "epoch": 19.5, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 24414 + }, + { + "epoch": 19.500798722044728, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 24415 + }, + { + "epoch": 19.501597444089455, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24416 + }, + { + "epoch": 19.502396166134186, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 24417 + }, + { + "epoch": 19.503194888178914, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 24418 + }, + { + "epoch": 19.50399361022364, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 24419 + }, + { + "epoch": 19.50479233226837, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 24420 + }, + { + "epoch": 19.5055910543131, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 24421 + }, + { + "epoch": 19.50638977635783, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 24422 + }, + { + "epoch": 19.507188498402556, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 24423 + }, + { + "epoch": 19.507987220447284, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 24424 + }, + { + "epoch": 19.50878594249201, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0701, + "step": 24425 + }, + { + "epoch": 19.509584664536742, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 24426 + }, + { + "epoch": 19.51038338658147, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 24427 + }, + { + "epoch": 19.511182108626198, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24428 + }, + { + "epoch": 19.511980830670925, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 24429 + }, + { + "epoch": 19.512779552715656, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 24430 + }, + { + "epoch": 19.513578274760384, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 24431 + }, + { + "epoch": 19.51437699680511, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 24432 + }, + { + "epoch": 19.51517571884984, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 24433 + }, + { + "epoch": 19.515974440894567, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 24434 + }, + { + "epoch": 19.516773162939298, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 24435 + }, + { + "epoch": 19.517571884984026, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 24436 + }, + { + "epoch": 19.518370607028753, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 24437 + }, + { + "epoch": 19.51916932907348, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 24438 + }, + { + "epoch": 19.519968051118212, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 24439 + }, + { + "epoch": 19.52076677316294, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24440 + }, + { + "epoch": 19.521565495207668, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 24441 + }, + { + "epoch": 19.522364217252395, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 24442 + }, + { + "epoch": 19.523162939297123, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 24443 + }, + { + "epoch": 19.523961661341854, + "grad_norm": 0.337890625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 24444 + }, + { + "epoch": 19.52476038338658, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 24445 + }, + { + "epoch": 19.52555910543131, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 24446 + }, + { + "epoch": 19.526357827476037, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 24447 + }, + { + "epoch": 19.527156549520768, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 24448 + }, + { + "epoch": 19.527955271565496, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 24449 + }, + { + "epoch": 19.528753993610223, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 24450 + }, + { + "epoch": 19.52955271565495, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 24451 + }, + { + "epoch": 19.53035143769968, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 24452 + }, + { + "epoch": 19.53115015974441, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 24453 + }, + { + "epoch": 19.531948881789138, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 24454 + }, + { + "epoch": 19.532747603833865, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 24455 + }, + { + "epoch": 19.533546325878593, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 24456 + }, + { + "epoch": 19.534345047923324, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 24457 + }, + { + "epoch": 19.53514376996805, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24458 + }, + { + "epoch": 19.53594249201278, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 24459 + }, + { + "epoch": 19.536741214057507, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 24460 + }, + { + "epoch": 19.537539936102235, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 24461 + }, + { + "epoch": 19.538338658146966, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 24462 + }, + { + "epoch": 19.539137380191693, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 24463 + }, + { + "epoch": 19.53993610223642, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 24464 + }, + { + "epoch": 19.54073482428115, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 24465 + }, + { + "epoch": 19.54153354632588, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24466 + }, + { + "epoch": 19.542332268370608, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24467 + }, + { + "epoch": 19.543130990415335, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 24468 + }, + { + "epoch": 19.543929712460063, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 24469 + }, + { + "epoch": 19.544728434504794, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 24470 + }, + { + "epoch": 19.54552715654952, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 24471 + }, + { + "epoch": 19.54632587859425, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 24472 + }, + { + "epoch": 19.547124600638977, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 24473 + }, + { + "epoch": 19.547923322683705, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 24474 + }, + { + "epoch": 19.548722044728436, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 24475 + }, + { + "epoch": 19.549520766773163, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 24476 + }, + { + "epoch": 19.55031948881789, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 24477 + }, + { + "epoch": 19.55111821086262, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24478 + }, + { + "epoch": 19.55191693290735, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 24479 + }, + { + "epoch": 19.552715654952078, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 24480 + }, + { + "epoch": 19.553514376996805, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 24481 + }, + { + "epoch": 19.554313099041533, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 24482 + }, + { + "epoch": 19.55511182108626, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 24483 + }, + { + "epoch": 19.55591054313099, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 24484 + }, + { + "epoch": 19.55670926517572, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 24485 + }, + { + "epoch": 19.557507987220447, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 24486 + }, + { + "epoch": 19.558306709265175, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24487 + }, + { + "epoch": 19.559105431309906, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 24488 + }, + { + "epoch": 19.559904153354633, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 24489 + }, + { + "epoch": 19.56070287539936, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 24490 + }, + { + "epoch": 19.56150159744409, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 24491 + }, + { + "epoch": 19.562300319488816, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 24492 + }, + { + "epoch": 19.563099041533548, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 24493 + }, + { + "epoch": 19.563897763578275, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 24494 + }, + { + "epoch": 19.564696485623003, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 24495 + }, + { + "epoch": 19.56549520766773, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 24496 + }, + { + "epoch": 19.56629392971246, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 24497 + }, + { + "epoch": 19.56709265175719, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24498 + }, + { + "epoch": 19.567891373801917, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 24499 + }, + { + "epoch": 19.568690095846645, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24500 + }, + { + "epoch": 19.569488817891372, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 24501 + }, + { + "epoch": 19.570287539936103, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 24502 + }, + { + "epoch": 19.57108626198083, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 24503 + }, + { + "epoch": 19.57188498402556, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 24504 + }, + { + "epoch": 19.572683706070286, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24505 + }, + { + "epoch": 19.573482428115017, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 24506 + }, + { + "epoch": 19.574281150159745, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 24507 + }, + { + "epoch": 19.575079872204473, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 24508 + }, + { + "epoch": 19.5758785942492, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 24509 + }, + { + "epoch": 19.576677316293928, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 24510 + }, + { + "epoch": 19.57747603833866, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 24511 + }, + { + "epoch": 19.578274760383387, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 24512 + }, + { + "epoch": 19.579073482428115, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 24513 + }, + { + "epoch": 19.579872204472842, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 24514 + }, + { + "epoch": 19.580670926517573, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 24515 + }, + { + "epoch": 19.5814696485623, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 24516 + }, + { + "epoch": 19.58226837060703, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 24517 + }, + { + "epoch": 19.583067092651756, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 24518 + }, + { + "epoch": 19.583865814696484, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 24519 + }, + { + "epoch": 19.584664536741215, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 24520 + }, + { + "epoch": 19.585463258785943, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 24521 + }, + { + "epoch": 19.58626198083067, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 24522 + }, + { + "epoch": 19.587060702875398, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 24523 + }, + { + "epoch": 19.58785942492013, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 24524 + }, + { + "epoch": 19.588658146964857, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 24525 + }, + { + "epoch": 19.589456869009584, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 24526 + }, + { + "epoch": 19.590255591054312, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 24527 + }, + { + "epoch": 19.591054313099043, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 24528 + }, + { + "epoch": 19.59185303514377, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 24529 + }, + { + "epoch": 19.5926517571885, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 24530 + }, + { + "epoch": 19.593450479233226, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 24531 + }, + { + "epoch": 19.594249201277954, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 24532 + }, + { + "epoch": 19.595047923322685, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 24533 + }, + { + "epoch": 19.595846645367413, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 24534 + }, + { + "epoch": 19.59664536741214, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 24535 + }, + { + "epoch": 19.597444089456868, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24536 + }, + { + "epoch": 19.5982428115016, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 24537 + }, + { + "epoch": 19.599041533546327, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 24538 + }, + { + "epoch": 19.599840255591054, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 24539 + }, + { + "epoch": 19.600638977635782, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 24540 + }, + { + "epoch": 19.60143769968051, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 24541 + }, + { + "epoch": 19.60223642172524, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 24542 + }, + { + "epoch": 19.60303514376997, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 24543 + }, + { + "epoch": 19.603833865814696, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 24544 + }, + { + "epoch": 19.604632587859424, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 24545 + }, + { + "epoch": 19.605431309904155, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 24546 + }, + { + "epoch": 19.606230031948883, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 24547 + }, + { + "epoch": 19.60702875399361, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 24548 + }, + { + "epoch": 19.607827476038338, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 24549 + }, + { + "epoch": 19.608626198083066, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24550 + }, + { + "epoch": 19.609424920127797, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 24551 + }, + { + "epoch": 19.610223642172524, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 24552 + }, + { + "epoch": 19.611022364217252, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 24553 + }, + { + "epoch": 19.61182108626198, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 24554 + }, + { + "epoch": 19.61261980830671, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 24555 + }, + { + "epoch": 19.61341853035144, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 24556 + }, + { + "epoch": 19.614217252396166, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 24557 + }, + { + "epoch": 19.615015974440894, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 24558 + }, + { + "epoch": 19.61581469648562, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 24559 + }, + { + "epoch": 19.616613418530353, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 24560 + }, + { + "epoch": 19.61741214057508, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 24561 + }, + { + "epoch": 19.618210862619808, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 24562 + }, + { + "epoch": 19.619009584664536, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 24563 + }, + { + "epoch": 19.619808306709267, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 24564 + }, + { + "epoch": 19.620607028753994, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 24565 + }, + { + "epoch": 19.621405750798722, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 24566 + }, + { + "epoch": 19.62220447284345, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 24567 + }, + { + "epoch": 19.623003194888177, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 24568 + }, + { + "epoch": 19.62380191693291, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 24569 + }, + { + "epoch": 19.624600638977636, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 24570 + }, + { + "epoch": 19.625399361022364, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 24571 + }, + { + "epoch": 19.62619808306709, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 24572 + }, + { + "epoch": 19.626996805111823, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 24573 + }, + { + "epoch": 19.62779552715655, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24574 + }, + { + "epoch": 19.628594249201278, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 24575 + }, + { + "epoch": 19.629392971246006, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 24576 + }, + { + "epoch": 19.630191693290733, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 24577 + }, + { + "epoch": 19.630990415335464, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 24578 + }, + { + "epoch": 19.631789137380192, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 24579 + }, + { + "epoch": 19.63258785942492, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 24580 + }, + { + "epoch": 19.633386581469647, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24581 + }, + { + "epoch": 19.63418530351438, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 24582 + }, + { + "epoch": 19.634984025559106, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 24583 + }, + { + "epoch": 19.635782747603834, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 24584 + }, + { + "epoch": 19.63658146964856, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 24585 + }, + { + "epoch": 19.63738019169329, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 24586 + }, + { + "epoch": 19.63817891373802, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 24587 + }, + { + "epoch": 19.638977635782748, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 24588 + }, + { + "epoch": 19.639776357827476, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 24589 + }, + { + "epoch": 19.640575079872203, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 24590 + }, + { + "epoch": 19.641373801916934, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 24591 + }, + { + "epoch": 19.642172523961662, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 24592 + }, + { + "epoch": 19.64297124600639, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 24593 + }, + { + "epoch": 19.643769968051117, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 24594 + }, + { + "epoch": 19.644568690095845, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 24595 + }, + { + "epoch": 19.645367412140576, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24596 + }, + { + "epoch": 19.646166134185304, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 24597 + }, + { + "epoch": 19.64696485623003, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 24598 + }, + { + "epoch": 19.64776357827476, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 24599 + }, + { + "epoch": 19.64856230031949, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 24600 + }, + { + "epoch": 19.649361022364218, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 24601 + }, + { + "epoch": 19.650159744408946, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 24602 + }, + { + "epoch": 19.650958466453673, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 24603 + }, + { + "epoch": 19.6517571884984, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 24604 + }, + { + "epoch": 19.652555910543132, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 24605 + }, + { + "epoch": 19.65335463258786, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 24606 + }, + { + "epoch": 19.654153354632587, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 24607 + }, + { + "epoch": 19.654952076677315, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 24608 + }, + { + "epoch": 19.655750798722046, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 24609 + }, + { + "epoch": 19.656549520766774, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 24610 + }, + { + "epoch": 19.6573482428115, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 24611 + }, + { + "epoch": 19.65814696485623, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 24612 + }, + { + "epoch": 19.658945686900957, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 24613 + }, + { + "epoch": 19.659744408945688, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 24614 + }, + { + "epoch": 19.660543130990416, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 24615 + }, + { + "epoch": 19.661341853035143, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 24616 + }, + { + "epoch": 19.66214057507987, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24617 + }, + { + "epoch": 19.662939297124602, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 24618 + }, + { + "epoch": 19.66373801916933, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 24619 + }, + { + "epoch": 19.664536741214057, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 24620 + }, + { + "epoch": 19.665335463258785, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 24621 + }, + { + "epoch": 19.666134185303516, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24622 + }, + { + "epoch": 19.666932907348244, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 24623 + }, + { + "epoch": 19.66773162939297, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 24624 + }, + { + "epoch": 19.6685303514377, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 24625 + }, + { + "epoch": 19.669329073482427, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 24626 + }, + { + "epoch": 19.670127795527158, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 24627 + }, + { + "epoch": 19.670926517571885, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 24628 + }, + { + "epoch": 19.671725239616613, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24629 + }, + { + "epoch": 19.67252396166134, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 24630 + }, + { + "epoch": 19.673322683706072, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 24631 + }, + { + "epoch": 19.6741214057508, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24632 + }, + { + "epoch": 19.674920127795527, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 24633 + }, + { + "epoch": 19.675718849840255, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 24634 + }, + { + "epoch": 19.676517571884983, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24635 + }, + { + "epoch": 19.677316293929714, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 24636 + }, + { + "epoch": 19.67811501597444, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 24637 + }, + { + "epoch": 19.67891373801917, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 24638 + }, + { + "epoch": 19.679712460063897, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 24639 + }, + { + "epoch": 19.680511182108628, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 24640 + }, + { + "epoch": 19.681309904153355, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 24641 + }, + { + "epoch": 19.682108626198083, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24642 + }, + { + "epoch": 19.68290734824281, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 24643 + }, + { + "epoch": 19.68370607028754, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 24644 + }, + { + "epoch": 19.68450479233227, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24645 + }, + { + "epoch": 19.685303514376997, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 24646 + }, + { + "epoch": 19.686102236421725, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 24647 + }, + { + "epoch": 19.686900958466452, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 24648 + }, + { + "epoch": 19.687699680511184, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 24649 + }, + { + "epoch": 19.68849840255591, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 24650 + }, + { + "epoch": 19.68929712460064, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 24651 + }, + { + "epoch": 19.690095846645367, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 24652 + }, + { + "epoch": 19.690894568690094, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 24653 + }, + { + "epoch": 19.691693290734825, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 24654 + }, + { + "epoch": 19.692492012779553, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 24655 + }, + { + "epoch": 19.69329073482428, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 24656 + }, + { + "epoch": 19.69408945686901, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 24657 + }, + { + "epoch": 19.69488817891374, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 24658 + }, + { + "epoch": 19.695686900958467, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 24659 + }, + { + "epoch": 19.696485623003195, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 24660 + }, + { + "epoch": 19.697284345047922, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 24661 + }, + { + "epoch": 19.69808306709265, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 24662 + }, + { + "epoch": 19.69888178913738, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 24663 + }, + { + "epoch": 19.69968051118211, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 24664 + }, + { + "epoch": 19.700479233226837, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24665 + }, + { + "epoch": 19.701277955271564, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 24666 + }, + { + "epoch": 19.702076677316295, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24667 + }, + { + "epoch": 19.702875399361023, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 24668 + }, + { + "epoch": 19.70367412140575, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 24669 + }, + { + "epoch": 19.70447284345048, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 24670 + }, + { + "epoch": 19.705271565495206, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 24671 + }, + { + "epoch": 19.706070287539937, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 24672 + }, + { + "epoch": 19.706869009584665, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 24673 + }, + { + "epoch": 19.707667731629392, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 24674 + }, + { + "epoch": 19.70846645367412, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 24675 + }, + { + "epoch": 19.70926517571885, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 24676 + }, + { + "epoch": 19.71006389776358, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 24677 + }, + { + "epoch": 19.710862619808307, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 24678 + }, + { + "epoch": 19.711661341853034, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 24679 + }, + { + "epoch": 19.712460063897765, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 24680 + }, + { + "epoch": 19.713258785942493, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 24681 + }, + { + "epoch": 19.71405750798722, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 24682 + }, + { + "epoch": 19.71485623003195, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 24683 + }, + { + "epoch": 19.715654952076676, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24684 + }, + { + "epoch": 19.716453674121407, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 24685 + }, + { + "epoch": 19.717252396166135, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 24686 + }, + { + "epoch": 19.718051118210862, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 24687 + }, + { + "epoch": 19.71884984025559, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 24688 + }, + { + "epoch": 19.71964856230032, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 24689 + }, + { + "epoch": 19.72044728434505, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 24690 + }, + { + "epoch": 19.721246006389777, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 24691 + }, + { + "epoch": 19.722044728434504, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 24692 + }, + { + "epoch": 19.722843450479232, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 24693 + }, + { + "epoch": 19.723642172523963, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 24694 + }, + { + "epoch": 19.72444089456869, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 24695 + }, + { + "epoch": 19.72523961661342, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 24696 + }, + { + "epoch": 19.726038338658146, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 24697 + }, + { + "epoch": 19.726837060702877, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24698 + }, + { + "epoch": 19.727635782747605, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 24699 + }, + { + "epoch": 19.728434504792332, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 24700 + }, + { + "epoch": 19.72923322683706, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 24701 + }, + { + "epoch": 19.730031948881788, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 24702 + }, + { + "epoch": 19.73083067092652, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 24703 + }, + { + "epoch": 19.731629392971247, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 24704 + }, + { + "epoch": 19.732428115015974, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 24705 + }, + { + "epoch": 19.733226837060702, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 24706 + }, + { + "epoch": 19.734025559105433, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24707 + }, + { + "epoch": 19.73482428115016, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24708 + }, + { + "epoch": 19.73562300319489, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 24709 + }, + { + "epoch": 19.736421725239616, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24710 + }, + { + "epoch": 19.737220447284344, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 24711 + }, + { + "epoch": 19.738019169329075, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 24712 + }, + { + "epoch": 19.738817891373802, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 24713 + }, + { + "epoch": 19.73961661341853, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 24714 + }, + { + "epoch": 19.740415335463258, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 24715 + }, + { + "epoch": 19.74121405750799, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 24716 + }, + { + "epoch": 19.742012779552716, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 24717 + }, + { + "epoch": 19.742811501597444, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 24718 + }, + { + "epoch": 19.74361022364217, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24719 + }, + { + "epoch": 19.7444089456869, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 24720 + }, + { + "epoch": 19.74520766773163, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24721 + }, + { + "epoch": 19.74600638977636, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 24722 + }, + { + "epoch": 19.746805111821086, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 24723 + }, + { + "epoch": 19.747603833865814, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 24724 + }, + { + "epoch": 19.748402555910545, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24725 + }, + { + "epoch": 19.749201277955272, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 24726 + }, + { + "epoch": 19.75, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 24727 + }, + { + "epoch": 19.750798722044728, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 24728 + }, + { + "epoch": 19.751597444089455, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 24729 + }, + { + "epoch": 19.752396166134186, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 24730 + }, + { + "epoch": 19.753194888178914, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 24731 + }, + { + "epoch": 19.75399361022364, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 24732 + }, + { + "epoch": 19.75479233226837, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 24733 + }, + { + "epoch": 19.7555910543131, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 24734 + }, + { + "epoch": 19.75638977635783, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 24735 + }, + { + "epoch": 19.757188498402556, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24736 + }, + { + "epoch": 19.757987220447284, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 24737 + }, + { + "epoch": 19.75878594249201, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24738 + }, + { + "epoch": 19.759584664536742, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 24739 + }, + { + "epoch": 19.76038338658147, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 24740 + }, + { + "epoch": 19.761182108626198, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 24741 + }, + { + "epoch": 19.761980830670925, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24742 + }, + { + "epoch": 19.762779552715656, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 24743 + }, + { + "epoch": 19.763578274760384, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 24744 + }, + { + "epoch": 19.76437699680511, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 24745 + }, + { + "epoch": 19.76517571884984, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24746 + }, + { + "epoch": 19.765974440894567, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 24747 + }, + { + "epoch": 19.766773162939298, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 24748 + }, + { + "epoch": 19.767571884984026, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 24749 + }, + { + "epoch": 19.768370607028753, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 24750 + }, + { + "epoch": 19.76916932907348, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 24751 + }, + { + "epoch": 19.769968051118212, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 24752 + }, + { + "epoch": 19.77076677316294, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 24753 + }, + { + "epoch": 19.771565495207668, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 24754 + }, + { + "epoch": 19.772364217252395, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 24755 + }, + { + "epoch": 19.773162939297123, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 24756 + }, + { + "epoch": 19.773961661341854, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 24757 + }, + { + "epoch": 19.77476038338658, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 24758 + }, + { + "epoch": 19.77555910543131, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 24759 + }, + { + "epoch": 19.776357827476037, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 24760 + }, + { + "epoch": 19.777156549520768, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 24761 + }, + { + "epoch": 19.777955271565496, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 24762 + }, + { + "epoch": 19.778753993610223, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 24763 + }, + { + "epoch": 19.77955271565495, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 24764 + }, + { + "epoch": 19.78035143769968, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 24765 + }, + { + "epoch": 19.78115015974441, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 24766 + }, + { + "epoch": 19.781948881789138, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24767 + }, + { + "epoch": 19.782747603833865, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 24768 + }, + { + "epoch": 19.783546325878593, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 24769 + }, + { + "epoch": 19.784345047923324, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 24770 + }, + { + "epoch": 19.78514376996805, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 24771 + }, + { + "epoch": 19.78594249201278, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 24772 + }, + { + "epoch": 19.786741214057507, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 24773 + }, + { + "epoch": 19.787539936102235, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 24774 + }, + { + "epoch": 19.788338658146966, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 24775 + }, + { + "epoch": 19.789137380191693, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 24776 + }, + { + "epoch": 19.78993610223642, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 24777 + }, + { + "epoch": 19.79073482428115, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 24778 + }, + { + "epoch": 19.79153354632588, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 24779 + }, + { + "epoch": 19.792332268370608, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24780 + }, + { + "epoch": 19.793130990415335, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 24781 + }, + { + "epoch": 19.793929712460063, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 24782 + }, + { + "epoch": 19.794728434504794, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 24783 + }, + { + "epoch": 19.79552715654952, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 24784 + }, + { + "epoch": 19.79632587859425, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 24785 + }, + { + "epoch": 19.797124600638977, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 24786 + }, + { + "epoch": 19.797923322683705, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 24787 + }, + { + "epoch": 19.798722044728436, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 24788 + }, + { + "epoch": 19.799520766773163, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 24789 + }, + { + "epoch": 19.80031948881789, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 24790 + }, + { + "epoch": 19.80111821086262, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 24791 + }, + { + "epoch": 19.80191693290735, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24792 + }, + { + "epoch": 19.802715654952078, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 24793 + }, + { + "epoch": 19.803514376996805, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 24794 + }, + { + "epoch": 19.804313099041533, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 24795 + }, + { + "epoch": 19.80511182108626, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 24796 + }, + { + "epoch": 19.80591054313099, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 24797 + }, + { + "epoch": 19.80670926517572, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 24798 + }, + { + "epoch": 19.807507987220447, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24799 + }, + { + "epoch": 19.808306709265175, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 24800 + }, + { + "epoch": 19.809105431309906, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 24801 + }, + { + "epoch": 19.809904153354633, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24802 + }, + { + "epoch": 19.81070287539936, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 24803 + }, + { + "epoch": 19.81150159744409, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 24804 + }, + { + "epoch": 19.812300319488816, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 24805 + }, + { + "epoch": 19.813099041533548, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 24806 + }, + { + "epoch": 19.813897763578275, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 24807 + }, + { + "epoch": 19.814696485623003, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 24808 + }, + { + "epoch": 19.81549520766773, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 24809 + }, + { + "epoch": 19.81629392971246, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 24810 + }, + { + "epoch": 19.81709265175719, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 24811 + }, + { + "epoch": 19.817891373801917, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 24812 + }, + { + "epoch": 19.818690095846645, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 24813 + }, + { + "epoch": 19.819488817891372, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24814 + }, + { + "epoch": 19.820287539936103, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 24815 + }, + { + "epoch": 19.82108626198083, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 24816 + }, + { + "epoch": 19.82188498402556, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 24817 + }, + { + "epoch": 19.822683706070286, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24818 + }, + { + "epoch": 19.823482428115017, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 24819 + }, + { + "epoch": 19.824281150159745, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 24820 + }, + { + "epoch": 19.825079872204473, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 24821 + }, + { + "epoch": 19.8258785942492, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 24822 + }, + { + "epoch": 19.826677316293928, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 24823 + }, + { + "epoch": 19.82747603833866, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 24824 + }, + { + "epoch": 19.828274760383387, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 24825 + }, + { + "epoch": 19.829073482428115, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 24826 + }, + { + "epoch": 19.829872204472842, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 24827 + }, + { + "epoch": 19.830670926517573, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 24828 + }, + { + "epoch": 19.8314696485623, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24829 + }, + { + "epoch": 19.83226837060703, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 24830 + }, + { + "epoch": 19.833067092651756, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 24831 + }, + { + "epoch": 19.833865814696484, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 24832 + }, + { + "epoch": 19.834664536741215, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 24833 + }, + { + "epoch": 19.835463258785943, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 24834 + }, + { + "epoch": 19.83626198083067, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 24835 + }, + { + "epoch": 19.837060702875398, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 24836 + }, + { + "epoch": 19.83785942492013, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 24837 + }, + { + "epoch": 19.838658146964857, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1078, + "step": 24838 + }, + { + "epoch": 19.839456869009584, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 24839 + }, + { + "epoch": 19.840255591054312, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 24840 + }, + { + "epoch": 19.841054313099043, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 24841 + }, + { + "epoch": 19.84185303514377, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 24842 + }, + { + "epoch": 19.8426517571885, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 24843 + }, + { + "epoch": 19.843450479233226, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 24844 + }, + { + "epoch": 19.844249201277954, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 24845 + }, + { + "epoch": 19.845047923322685, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 24846 + }, + { + "epoch": 19.845846645367413, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 24847 + }, + { + "epoch": 19.84664536741214, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 24848 + }, + { + "epoch": 19.847444089456868, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 24849 + }, + { + "epoch": 19.8482428115016, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 24850 + }, + { + "epoch": 19.849041533546327, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 24851 + }, + { + "epoch": 19.849840255591054, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 24852 + }, + { + "epoch": 19.850638977635782, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 24853 + }, + { + "epoch": 19.85143769968051, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 24854 + }, + { + "epoch": 19.85223642172524, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 24855 + }, + { + "epoch": 19.85303514376997, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 24856 + }, + { + "epoch": 19.853833865814696, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 24857 + }, + { + "epoch": 19.854632587859424, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 24858 + }, + { + "epoch": 19.855431309904155, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 24859 + }, + { + "epoch": 19.856230031948883, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 24860 + }, + { + "epoch": 19.85702875399361, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 24861 + }, + { + "epoch": 19.857827476038338, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 24862 + }, + { + "epoch": 19.858626198083066, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24863 + }, + { + "epoch": 19.859424920127797, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 24864 + }, + { + "epoch": 19.860223642172524, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 24865 + }, + { + "epoch": 19.861022364217252, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 24866 + }, + { + "epoch": 19.86182108626198, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 24867 + }, + { + "epoch": 19.86261980830671, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 24868 + }, + { + "epoch": 19.86341853035144, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 24869 + }, + { + "epoch": 19.864217252396166, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 24870 + }, + { + "epoch": 19.865015974440894, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 24871 + }, + { + "epoch": 19.86581469648562, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 24872 + }, + { + "epoch": 19.866613418530353, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 24873 + }, + { + "epoch": 19.86741214057508, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 24874 + }, + { + "epoch": 19.868210862619808, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 24875 + }, + { + "epoch": 19.869009584664536, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 24876 + }, + { + "epoch": 19.869808306709267, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 24877 + }, + { + "epoch": 19.870607028753994, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 24878 + }, + { + "epoch": 19.871405750798722, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 24879 + }, + { + "epoch": 19.87220447284345, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 24880 + }, + { + "epoch": 19.873003194888177, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 24881 + }, + { + "epoch": 19.87380191693291, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 24882 + }, + { + "epoch": 19.874600638977636, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 24883 + }, + { + "epoch": 19.875399361022364, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 24884 + }, + { + "epoch": 19.87619808306709, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24885 + }, + { + "epoch": 19.876996805111823, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 24886 + }, + { + "epoch": 19.87779552715655, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 24887 + }, + { + "epoch": 19.878594249201278, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 24888 + }, + { + "epoch": 19.879392971246006, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 24889 + }, + { + "epoch": 19.880191693290733, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 24890 + }, + { + "epoch": 19.880990415335464, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 24891 + }, + { + "epoch": 19.881789137380192, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 24892 + }, + { + "epoch": 19.88258785942492, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 24893 + }, + { + "epoch": 19.883386581469647, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 24894 + }, + { + "epoch": 19.88418530351438, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 24895 + }, + { + "epoch": 19.884984025559106, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 24896 + }, + { + "epoch": 19.885782747603834, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 24897 + }, + { + "epoch": 19.88658146964856, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 24898 + }, + { + "epoch": 19.88738019169329, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 24899 + }, + { + "epoch": 19.88817891373802, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 24900 + }, + { + "epoch": 19.888977635782748, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 24901 + }, + { + "epoch": 19.889776357827476, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24902 + }, + { + "epoch": 19.890575079872203, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 24903 + }, + { + "epoch": 19.891373801916934, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 24904 + }, + { + "epoch": 19.892172523961662, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 24905 + }, + { + "epoch": 19.89297124600639, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 24906 + }, + { + "epoch": 19.893769968051117, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 24907 + }, + { + "epoch": 19.894568690095845, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 24908 + }, + { + "epoch": 19.895367412140576, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 24909 + }, + { + "epoch": 19.896166134185304, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 24910 + }, + { + "epoch": 19.89696485623003, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 24911 + }, + { + "epoch": 19.89776357827476, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 24912 + }, + { + "epoch": 19.89856230031949, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 24913 + }, + { + "epoch": 19.899361022364218, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 24914 + }, + { + "epoch": 19.900159744408946, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 24915 + }, + { + "epoch": 19.900958466453673, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 24916 + }, + { + "epoch": 19.9017571884984, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 24917 + }, + { + "epoch": 19.902555910543132, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 24918 + }, + { + "epoch": 19.90335463258786, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 24919 + }, + { + "epoch": 19.904153354632587, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 24920 + }, + { + "epoch": 19.904952076677315, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 24921 + }, + { + "epoch": 19.905750798722046, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 24922 + }, + { + "epoch": 19.906549520766774, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 24923 + }, + { + "epoch": 19.9073482428115, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 24924 + }, + { + "epoch": 19.90814696485623, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 24925 + }, + { + "epoch": 19.908945686900957, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1084, + "step": 24926 + }, + { + "epoch": 19.909744408945688, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 24927 + }, + { + "epoch": 19.910543130990416, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 24928 + }, + { + "epoch": 19.911341853035143, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24929 + }, + { + "epoch": 19.91214057507987, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 24930 + }, + { + "epoch": 19.912939297124602, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 24931 + }, + { + "epoch": 19.91373801916933, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 24932 + }, + { + "epoch": 19.914536741214057, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 24933 + }, + { + "epoch": 19.915335463258785, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 24934 + }, + { + "epoch": 19.916134185303516, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 24935 + }, + { + "epoch": 19.916932907348244, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 24936 + }, + { + "epoch": 19.91773162939297, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1058, + "step": 24937 + }, + { + "epoch": 19.9185303514377, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 24938 + }, + { + "epoch": 19.919329073482427, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 24939 + }, + { + "epoch": 19.920127795527158, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 24940 + }, + { + "epoch": 19.920926517571885, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 24941 + }, + { + "epoch": 19.921725239616613, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 24942 + }, + { + "epoch": 19.92252396166134, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 24943 + }, + { + "epoch": 19.923322683706072, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 24944 + }, + { + "epoch": 19.9241214057508, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 24945 + }, + { + "epoch": 19.924920127795527, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 24946 + }, + { + "epoch": 19.925718849840255, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 24947 + }, + { + "epoch": 19.926517571884983, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 24948 + }, + { + "epoch": 19.927316293929714, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 24949 + }, + { + "epoch": 19.92811501597444, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 24950 + }, + { + "epoch": 19.92891373801917, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 24951 + }, + { + "epoch": 19.929712460063897, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 24952 + }, + { + "epoch": 19.930511182108628, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 24953 + }, + { + "epoch": 19.931309904153355, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 24954 + }, + { + "epoch": 19.932108626198083, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 24955 + }, + { + "epoch": 19.93290734824281, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 24956 + }, + { + "epoch": 19.93370607028754, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 24957 + }, + { + "epoch": 19.93450479233227, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 24958 + }, + { + "epoch": 19.935303514376997, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 24959 + }, + { + "epoch": 19.936102236421725, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 24960 + }, + { + "epoch": 19.936900958466452, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 24961 + }, + { + "epoch": 19.937699680511184, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 24962 + }, + { + "epoch": 19.93849840255591, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 24963 + }, + { + "epoch": 19.93929712460064, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 24964 + }, + { + "epoch": 19.940095846645367, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 24965 + }, + { + "epoch": 19.940894568690094, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 24966 + }, + { + "epoch": 19.941693290734825, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 24967 + }, + { + "epoch": 19.942492012779553, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 24968 + }, + { + "epoch": 19.94329073482428, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 24969 + }, + { + "epoch": 19.94408945686901, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 24970 + }, + { + "epoch": 19.94488817891374, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 24971 + }, + { + "epoch": 19.945686900958467, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 24972 + }, + { + "epoch": 19.946485623003195, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 24973 + }, + { + "epoch": 19.947284345047922, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 24974 + }, + { + "epoch": 19.94808306709265, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 24975 + }, + { + "epoch": 19.94888178913738, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 24976 + }, + { + "epoch": 19.94968051118211, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 24977 + }, + { + "epoch": 19.950479233226837, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 24978 + }, + { + "epoch": 19.951277955271564, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 24979 + }, + { + "epoch": 19.952076677316295, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 24980 + }, + { + "epoch": 19.952875399361023, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 24981 + }, + { + "epoch": 19.95367412140575, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 24982 + }, + { + "epoch": 19.95447284345048, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 24983 + }, + { + "epoch": 19.955271565495206, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 24984 + }, + { + "epoch": 19.956070287539937, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 24985 + }, + { + "epoch": 19.956869009584665, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 24986 + }, + { + "epoch": 19.957667731629392, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 24987 + }, + { + "epoch": 19.95846645367412, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 24988 + }, + { + "epoch": 19.95926517571885, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 24989 + }, + { + "epoch": 19.96006389776358, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 24990 + }, + { + "epoch": 19.960862619808307, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 24991 + }, + { + "epoch": 19.961661341853034, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 24992 + }, + { + "epoch": 19.962460063897765, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 24993 + }, + { + "epoch": 19.963258785942493, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 24994 + }, + { + "epoch": 19.96405750798722, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 24995 + }, + { + "epoch": 19.96485623003195, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 24996 + }, + { + "epoch": 19.965654952076676, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 24997 + }, + { + "epoch": 19.966453674121407, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 24998 + }, + { + "epoch": 19.967252396166135, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 24999 + }, + { + "epoch": 19.968051118210862, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 25000 + }, + { + "epoch": 19.96884984025559, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 25001 + }, + { + "epoch": 19.96964856230032, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 25002 + }, + { + "epoch": 19.97044728434505, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 25003 + }, + { + "epoch": 19.971246006389777, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 25004 + }, + { + "epoch": 19.972044728434504, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 25005 + }, + { + "epoch": 19.972843450479232, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25006 + }, + { + "epoch": 19.973642172523963, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 25007 + }, + { + "epoch": 19.97444089456869, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 25008 + }, + { + "epoch": 19.97523961661342, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 25009 + }, + { + "epoch": 19.976038338658146, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 25010 + }, + { + "epoch": 19.976837060702877, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 25011 + }, + { + "epoch": 19.977635782747605, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 25012 + }, + { + "epoch": 19.978434504792332, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 25013 + }, + { + "epoch": 19.97923322683706, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 25014 + }, + { + "epoch": 19.980031948881788, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 25015 + }, + { + "epoch": 19.98083067092652, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 25016 + }, + { + "epoch": 19.981629392971247, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 25017 + }, + { + "epoch": 19.982428115015974, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 25018 + }, + { + "epoch": 19.983226837060702, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 25019 + }, + { + "epoch": 19.984025559105433, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 25020 + }, + { + "epoch": 19.98482428115016, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 25021 + }, + { + "epoch": 19.98562300319489, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 25022 + }, + { + "epoch": 19.986421725239616, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25023 + }, + { + "epoch": 19.987220447284344, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 25024 + }, + { + "epoch": 19.988019169329075, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 25025 + }, + { + "epoch": 19.988817891373802, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 25026 + }, + { + "epoch": 19.98961661341853, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 25027 + }, + { + "epoch": 19.990415335463258, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 25028 + }, + { + "epoch": 19.99121405750799, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 25029 + }, + { + "epoch": 19.992012779552716, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 25030 + }, + { + "epoch": 19.992811501597444, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 25031 + }, + { + "epoch": 19.99361022364217, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 25032 + }, + { + "epoch": 19.9944089456869, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25033 + }, + { + "epoch": 19.99520766773163, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 25034 + }, + { + "epoch": 19.99600638977636, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 25035 + }, + { + "epoch": 19.996805111821086, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 25036 + }, + { + "epoch": 19.997603833865814, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 25037 + }, + { + "epoch": 19.998402555910545, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 25038 + }, + { + "epoch": 19.999201277955272, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 25039 + }, + { + "epoch": 20.0, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 25040 + }, + { + "epoch": 20.000798722044728, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 25041 + }, + { + "epoch": 20.001597444089455, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 25042 + }, + { + "epoch": 20.002396166134186, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 25043 + }, + { + "epoch": 20.003194888178914, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25044 + }, + { + "epoch": 20.00399361022364, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 25045 + }, + { + "epoch": 20.00479233226837, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 25046 + }, + { + "epoch": 20.0055910543131, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 25047 + }, + { + "epoch": 20.00638977635783, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 25048 + }, + { + "epoch": 20.007188498402556, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25049 + }, + { + "epoch": 20.007987220447284, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 25050 + }, + { + "epoch": 20.00878594249201, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 25051 + }, + { + "epoch": 20.009584664536742, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 25052 + }, + { + "epoch": 20.01038338658147, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 25053 + }, + { + "epoch": 20.011182108626198, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 25054 + }, + { + "epoch": 20.011980830670925, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25055 + }, + { + "epoch": 20.012779552715656, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 25056 + }, + { + "epoch": 20.013578274760384, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 25057 + }, + { + "epoch": 20.01437699680511, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 25058 + }, + { + "epoch": 20.01517571884984, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 25059 + }, + { + "epoch": 20.015974440894567, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 25060 + }, + { + "epoch": 20.016773162939298, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 25061 + }, + { + "epoch": 20.017571884984026, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 25062 + }, + { + "epoch": 20.018370607028753, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 25063 + }, + { + "epoch": 20.01916932907348, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 25064 + }, + { + "epoch": 20.019968051118212, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25065 + }, + { + "epoch": 20.02076677316294, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 25066 + }, + { + "epoch": 20.021565495207668, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 25067 + }, + { + "epoch": 20.022364217252395, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 25068 + }, + { + "epoch": 20.023162939297123, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25069 + }, + { + "epoch": 20.023961661341854, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 25070 + }, + { + "epoch": 20.02476038338658, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 25071 + }, + { + "epoch": 20.02555910543131, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25072 + }, + { + "epoch": 20.026357827476037, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 25073 + }, + { + "epoch": 20.027156549520768, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 25074 + }, + { + "epoch": 20.027955271565496, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 25075 + }, + { + "epoch": 20.028753993610223, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 25076 + }, + { + "epoch": 20.02955271565495, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 25077 + }, + { + "epoch": 20.03035143769968, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 25078 + }, + { + "epoch": 20.03115015974441, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 25079 + }, + { + "epoch": 20.031948881789138, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 25080 + }, + { + "epoch": 20.032747603833865, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 25081 + }, + { + "epoch": 20.033546325878593, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 25082 + }, + { + "epoch": 20.034345047923324, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 25083 + }, + { + "epoch": 20.03514376996805, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 25084 + }, + { + "epoch": 20.03594249201278, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 25085 + }, + { + "epoch": 20.036741214057507, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 25086 + }, + { + "epoch": 20.037539936102238, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 25087 + }, + { + "epoch": 20.038338658146966, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 25088 + }, + { + "epoch": 20.039137380191693, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 25089 + }, + { + "epoch": 20.03993610223642, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 25090 + }, + { + "epoch": 20.04073482428115, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 25091 + }, + { + "epoch": 20.04153354632588, + "grad_norm": 0.0966796875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 25092 + }, + { + "epoch": 20.042332268370608, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 25093 + }, + { + "epoch": 20.043130990415335, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 25094 + }, + { + "epoch": 20.043929712460063, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25095 + }, + { + "epoch": 20.044728434504794, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 25096 + }, + { + "epoch": 20.04552715654952, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 25097 + }, + { + "epoch": 20.04632587859425, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 25098 + }, + { + "epoch": 20.047124600638977, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 25099 + }, + { + "epoch": 20.047923322683705, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 25100 + }, + { + "epoch": 20.048722044728436, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 25101 + }, + { + "epoch": 20.049520766773163, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25102 + }, + { + "epoch": 20.05031948881789, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 25103 + }, + { + "epoch": 20.05111821086262, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 25104 + }, + { + "epoch": 20.05191693290735, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 25105 + }, + { + "epoch": 20.052715654952078, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 25106 + }, + { + "epoch": 20.053514376996805, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 25107 + }, + { + "epoch": 20.054313099041533, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 25108 + }, + { + "epoch": 20.05511182108626, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 25109 + }, + { + "epoch": 20.05591054313099, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 25110 + }, + { + "epoch": 20.05670926517572, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 25111 + }, + { + "epoch": 20.057507987220447, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 25112 + }, + { + "epoch": 20.058306709265175, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 25113 + }, + { + "epoch": 20.059105431309906, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 25114 + }, + { + "epoch": 20.059904153354633, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 25115 + }, + { + "epoch": 20.06070287539936, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 25116 + }, + { + "epoch": 20.06150159744409, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 25117 + }, + { + "epoch": 20.062300319488816, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 25118 + }, + { + "epoch": 20.063099041533548, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 25119 + }, + { + "epoch": 20.063897763578275, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 25120 + }, + { + "epoch": 20.064696485623003, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 25121 + }, + { + "epoch": 20.06549520766773, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 25122 + }, + { + "epoch": 20.06629392971246, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 25123 + }, + { + "epoch": 20.06709265175719, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 25124 + }, + { + "epoch": 20.067891373801917, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 25125 + }, + { + "epoch": 20.068690095846645, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25126 + }, + { + "epoch": 20.069488817891372, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25127 + }, + { + "epoch": 20.070287539936103, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 25128 + }, + { + "epoch": 20.07108626198083, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25129 + }, + { + "epoch": 20.07188498402556, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 25130 + }, + { + "epoch": 20.072683706070286, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 25131 + }, + { + "epoch": 20.073482428115017, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 25132 + }, + { + "epoch": 20.074281150159745, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 25133 + }, + { + "epoch": 20.075079872204473, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1063, + "step": 25134 + }, + { + "epoch": 20.0758785942492, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 25135 + }, + { + "epoch": 20.076677316293928, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 25136 + }, + { + "epoch": 20.07747603833866, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 25137 + }, + { + "epoch": 20.078274760383387, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 25138 + }, + { + "epoch": 20.079073482428115, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 25139 + }, + { + "epoch": 20.079872204472842, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 25140 + }, + { + "epoch": 20.080670926517573, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 25141 + }, + { + "epoch": 20.0814696485623, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25142 + }, + { + "epoch": 20.08226837060703, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 25143 + }, + { + "epoch": 20.083067092651756, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 25144 + }, + { + "epoch": 20.083865814696484, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 25145 + }, + { + "epoch": 20.084664536741215, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 25146 + }, + { + "epoch": 20.085463258785943, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 25147 + }, + { + "epoch": 20.08626198083067, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 25148 + }, + { + "epoch": 20.087060702875398, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 25149 + }, + { + "epoch": 20.08785942492013, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 25150 + }, + { + "epoch": 20.088658146964857, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 25151 + }, + { + "epoch": 20.089456869009584, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 25152 + }, + { + "epoch": 20.090255591054312, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 25153 + }, + { + "epoch": 20.09105431309904, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 25154 + }, + { + "epoch": 20.09185303514377, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 25155 + }, + { + "epoch": 20.0926517571885, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 25156 + }, + { + "epoch": 20.093450479233226, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 25157 + }, + { + "epoch": 20.094249201277954, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 25158 + }, + { + "epoch": 20.095047923322685, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25159 + }, + { + "epoch": 20.095846645367413, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 25160 + }, + { + "epoch": 20.09664536741214, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 25161 + }, + { + "epoch": 20.097444089456868, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 25162 + }, + { + "epoch": 20.0982428115016, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 25163 + }, + { + "epoch": 20.099041533546327, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 25164 + }, + { + "epoch": 20.099840255591054, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 25165 + }, + { + "epoch": 20.100638977635782, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 25166 + }, + { + "epoch": 20.10143769968051, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 25167 + }, + { + "epoch": 20.10223642172524, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 25168 + }, + { + "epoch": 20.10303514376997, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25169 + }, + { + "epoch": 20.103833865814696, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 25170 + }, + { + "epoch": 20.104632587859424, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 25171 + }, + { + "epoch": 20.105431309904155, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 25172 + }, + { + "epoch": 20.106230031948883, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 25173 + }, + { + "epoch": 20.10702875399361, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 25174 + }, + { + "epoch": 20.107827476038338, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 25175 + }, + { + "epoch": 20.108626198083066, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 25176 + }, + { + "epoch": 20.109424920127797, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25177 + }, + { + "epoch": 20.110223642172524, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 25178 + }, + { + "epoch": 20.111022364217252, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 25179 + }, + { + "epoch": 20.11182108626198, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1071, + "step": 25180 + }, + { + "epoch": 20.11261980830671, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 25181 + }, + { + "epoch": 20.11341853035144, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 25182 + }, + { + "epoch": 20.114217252396166, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 25183 + }, + { + "epoch": 20.115015974440894, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 25184 + }, + { + "epoch": 20.11581469648562, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 25185 + }, + { + "epoch": 20.116613418530353, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 25186 + }, + { + "epoch": 20.11741214057508, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 25187 + }, + { + "epoch": 20.118210862619808, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 25188 + }, + { + "epoch": 20.119009584664536, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 25189 + }, + { + "epoch": 20.119808306709267, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 25190 + }, + { + "epoch": 20.120607028753994, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 25191 + }, + { + "epoch": 20.121405750798722, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 25192 + }, + { + "epoch": 20.12220447284345, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 25193 + }, + { + "epoch": 20.123003194888177, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 25194 + }, + { + "epoch": 20.12380191693291, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 25195 + }, + { + "epoch": 20.124600638977636, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 25196 + }, + { + "epoch": 20.125399361022364, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 25197 + }, + { + "epoch": 20.12619808306709, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 25198 + }, + { + "epoch": 20.126996805111823, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 25199 + }, + { + "epoch": 20.12779552715655, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25200 + }, + { + "epoch": 20.128594249201278, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 25201 + }, + { + "epoch": 20.129392971246006, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25202 + }, + { + "epoch": 20.130191693290733, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 25203 + }, + { + "epoch": 20.130990415335464, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 25204 + }, + { + "epoch": 20.131789137380192, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 25205 + }, + { + "epoch": 20.13258785942492, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25206 + }, + { + "epoch": 20.133386581469647, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 25207 + }, + { + "epoch": 20.13418530351438, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 25208 + }, + { + "epoch": 20.134984025559106, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 25209 + }, + { + "epoch": 20.135782747603834, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 25210 + }, + { + "epoch": 20.13658146964856, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 25211 + }, + { + "epoch": 20.13738019169329, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 25212 + }, + { + "epoch": 20.13817891373802, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 25213 + }, + { + "epoch": 20.138977635782748, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 25214 + }, + { + "epoch": 20.139776357827476, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25215 + }, + { + "epoch": 20.140575079872203, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 25216 + }, + { + "epoch": 20.141373801916934, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 25217 + }, + { + "epoch": 20.142172523961662, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 25218 + }, + { + "epoch": 20.14297124600639, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 25219 + }, + { + "epoch": 20.143769968051117, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 25220 + }, + { + "epoch": 20.144568690095845, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 25221 + }, + { + "epoch": 20.145367412140576, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 25222 + }, + { + "epoch": 20.146166134185304, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 25223 + }, + { + "epoch": 20.14696485623003, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 25224 + }, + { + "epoch": 20.14776357827476, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 25225 + }, + { + "epoch": 20.14856230031949, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 25226 + }, + { + "epoch": 20.149361022364218, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 25227 + }, + { + "epoch": 20.150159744408946, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 25228 + }, + { + "epoch": 20.150958466453673, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 25229 + }, + { + "epoch": 20.1517571884984, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 25230 + }, + { + "epoch": 20.152555910543132, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 25231 + }, + { + "epoch": 20.15335463258786, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 25232 + }, + { + "epoch": 20.154153354632587, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 25233 + }, + { + "epoch": 20.154952076677315, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 25234 + }, + { + "epoch": 20.155750798722046, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 25235 + }, + { + "epoch": 20.156549520766774, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 25236 + }, + { + "epoch": 20.1573482428115, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 25237 + }, + { + "epoch": 20.15814696485623, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 25238 + }, + { + "epoch": 20.15894568690096, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 25239 + }, + { + "epoch": 20.159744408945688, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 25240 + }, + { + "epoch": 20.160543130990416, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 25241 + }, + { + "epoch": 20.161341853035143, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25242 + }, + { + "epoch": 20.16214057507987, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25243 + }, + { + "epoch": 20.162939297124602, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 25244 + }, + { + "epoch": 20.16373801916933, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 25245 + }, + { + "epoch": 20.164536741214057, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 25246 + }, + { + "epoch": 20.165335463258785, + "grad_norm": 0.0947265625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 25247 + }, + { + "epoch": 20.166134185303516, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 25248 + }, + { + "epoch": 20.166932907348244, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 25249 + }, + { + "epoch": 20.16773162939297, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 25250 + }, + { + "epoch": 20.1685303514377, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 25251 + }, + { + "epoch": 20.169329073482427, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 25252 + }, + { + "epoch": 20.170127795527158, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 25253 + }, + { + "epoch": 20.170926517571885, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 25254 + }, + { + "epoch": 20.171725239616613, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 25255 + }, + { + "epoch": 20.17252396166134, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 25256 + }, + { + "epoch": 20.173322683706072, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 25257 + }, + { + "epoch": 20.1741214057508, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 25258 + }, + { + "epoch": 20.174920127795527, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 25259 + }, + { + "epoch": 20.175718849840255, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 25260 + }, + { + "epoch": 20.176517571884983, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 25261 + }, + { + "epoch": 20.177316293929714, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 25262 + }, + { + "epoch": 20.17811501597444, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 25263 + }, + { + "epoch": 20.17891373801917, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 25264 + }, + { + "epoch": 20.179712460063897, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 25265 + }, + { + "epoch": 20.180511182108628, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 25266 + }, + { + "epoch": 20.181309904153355, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 25267 + }, + { + "epoch": 20.182108626198083, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 25268 + }, + { + "epoch": 20.18290734824281, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 25269 + }, + { + "epoch": 20.18370607028754, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 25270 + }, + { + "epoch": 20.18450479233227, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 25271 + }, + { + "epoch": 20.185303514376997, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25272 + }, + { + "epoch": 20.186102236421725, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 25273 + }, + { + "epoch": 20.186900958466452, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 25274 + }, + { + "epoch": 20.187699680511184, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 25275 + }, + { + "epoch": 20.18849840255591, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 25276 + }, + { + "epoch": 20.18929712460064, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 25277 + }, + { + "epoch": 20.190095846645367, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 25278 + }, + { + "epoch": 20.190894568690094, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 25279 + }, + { + "epoch": 20.191693290734825, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 25280 + }, + { + "epoch": 20.192492012779553, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 25281 + }, + { + "epoch": 20.19329073482428, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 25282 + }, + { + "epoch": 20.19408945686901, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 25283 + }, + { + "epoch": 20.19488817891374, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 25284 + }, + { + "epoch": 20.195686900958467, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 25285 + }, + { + "epoch": 20.196485623003195, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 25286 + }, + { + "epoch": 20.197284345047922, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 25287 + }, + { + "epoch": 20.19808306709265, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 25288 + }, + { + "epoch": 20.19888178913738, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 25289 + }, + { + "epoch": 20.19968051118211, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 25290 + }, + { + "epoch": 20.200479233226837, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 25291 + }, + { + "epoch": 20.201277955271564, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 25292 + }, + { + "epoch": 20.202076677316295, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 25293 + }, + { + "epoch": 20.202875399361023, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 25294 + }, + { + "epoch": 20.20367412140575, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25295 + }, + { + "epoch": 20.20447284345048, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 25296 + }, + { + "epoch": 20.205271565495206, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 25297 + }, + { + "epoch": 20.206070287539937, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 25298 + }, + { + "epoch": 20.206869009584665, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 25299 + }, + { + "epoch": 20.207667731629392, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 25300 + }, + { + "epoch": 20.20846645367412, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 25301 + }, + { + "epoch": 20.20926517571885, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 25302 + }, + { + "epoch": 20.21006389776358, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 25303 + }, + { + "epoch": 20.210862619808307, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 25304 + }, + { + "epoch": 20.211661341853034, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 25305 + }, + { + "epoch": 20.212460063897762, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 25306 + }, + { + "epoch": 20.213258785942493, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 25307 + }, + { + "epoch": 20.21405750798722, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 25308 + }, + { + "epoch": 20.21485623003195, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 25309 + }, + { + "epoch": 20.215654952076676, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 25310 + }, + { + "epoch": 20.216453674121407, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 25311 + }, + { + "epoch": 20.217252396166135, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 25312 + }, + { + "epoch": 20.218051118210862, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 25313 + }, + { + "epoch": 20.21884984025559, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 25314 + }, + { + "epoch": 20.21964856230032, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25315 + }, + { + "epoch": 20.22044728434505, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 25316 + }, + { + "epoch": 20.221246006389777, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 25317 + }, + { + "epoch": 20.222044728434504, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 25318 + }, + { + "epoch": 20.222843450479232, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 25319 + }, + { + "epoch": 20.223642172523963, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 25320 + }, + { + "epoch": 20.22444089456869, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25321 + }, + { + "epoch": 20.22523961661342, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 25322 + }, + { + "epoch": 20.226038338658146, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 25323 + }, + { + "epoch": 20.226837060702877, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 25324 + }, + { + "epoch": 20.227635782747605, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 25325 + }, + { + "epoch": 20.228434504792332, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 25326 + }, + { + "epoch": 20.22923322683706, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 25327 + }, + { + "epoch": 20.230031948881788, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 25328 + }, + { + "epoch": 20.23083067092652, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 25329 + }, + { + "epoch": 20.231629392971247, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 25330 + }, + { + "epoch": 20.232428115015974, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 25331 + }, + { + "epoch": 20.233226837060702, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 25332 + }, + { + "epoch": 20.234025559105433, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 25333 + }, + { + "epoch": 20.23482428115016, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 25334 + }, + { + "epoch": 20.23562300319489, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 25335 + }, + { + "epoch": 20.236421725239616, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 25336 + }, + { + "epoch": 20.237220447284344, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 25337 + }, + { + "epoch": 20.238019169329075, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 25338 + }, + { + "epoch": 20.238817891373802, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 25339 + }, + { + "epoch": 20.23961661341853, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 25340 + }, + { + "epoch": 20.240415335463258, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25341 + }, + { + "epoch": 20.24121405750799, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 25342 + }, + { + "epoch": 20.242012779552716, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25343 + }, + { + "epoch": 20.242811501597444, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 25344 + }, + { + "epoch": 20.24361022364217, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 25345 + }, + { + "epoch": 20.2444089456869, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 25346 + }, + { + "epoch": 20.24520766773163, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 25347 + }, + { + "epoch": 20.24600638977636, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 25348 + }, + { + "epoch": 20.246805111821086, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 25349 + }, + { + "epoch": 20.247603833865814, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 25350 + }, + { + "epoch": 20.248402555910545, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 25351 + }, + { + "epoch": 20.249201277955272, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 25352 + }, + { + "epoch": 20.25, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 25353 + }, + { + "epoch": 20.250798722044728, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 25354 + }, + { + "epoch": 20.251597444089455, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 25355 + }, + { + "epoch": 20.252396166134186, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 25356 + }, + { + "epoch": 20.253194888178914, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 25357 + }, + { + "epoch": 20.25399361022364, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 25358 + }, + { + "epoch": 20.25479233226837, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 25359 + }, + { + "epoch": 20.2555910543131, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25360 + }, + { + "epoch": 20.25638977635783, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 25361 + }, + { + "epoch": 20.257188498402556, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 25362 + }, + { + "epoch": 20.257987220447284, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 25363 + }, + { + "epoch": 20.25878594249201, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 25364 + }, + { + "epoch": 20.259584664536742, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 25365 + }, + { + "epoch": 20.26038338658147, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 25366 + }, + { + "epoch": 20.261182108626198, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 25367 + }, + { + "epoch": 20.261980830670925, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 25368 + }, + { + "epoch": 20.262779552715656, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 25369 + }, + { + "epoch": 20.263578274760384, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 25370 + }, + { + "epoch": 20.26437699680511, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25371 + }, + { + "epoch": 20.26517571884984, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 25372 + }, + { + "epoch": 20.265974440894567, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 25373 + }, + { + "epoch": 20.266773162939298, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 25374 + }, + { + "epoch": 20.267571884984026, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 25375 + }, + { + "epoch": 20.268370607028753, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 25376 + }, + { + "epoch": 20.26916932907348, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 25377 + }, + { + "epoch": 20.269968051118212, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 25378 + }, + { + "epoch": 20.27076677316294, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 25379 + }, + { + "epoch": 20.271565495207668, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 25380 + }, + { + "epoch": 20.272364217252395, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 25381 + }, + { + "epoch": 20.273162939297123, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 25382 + }, + { + "epoch": 20.273961661341854, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25383 + }, + { + "epoch": 20.27476038338658, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 25384 + }, + { + "epoch": 20.27555910543131, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 25385 + }, + { + "epoch": 20.276357827476037, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 25386 + }, + { + "epoch": 20.277156549520768, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 25387 + }, + { + "epoch": 20.277955271565496, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 25388 + }, + { + "epoch": 20.278753993610223, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 25389 + }, + { + "epoch": 20.27955271565495, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 25390 + }, + { + "epoch": 20.28035143769968, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25391 + }, + { + "epoch": 20.28115015974441, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 25392 + }, + { + "epoch": 20.281948881789138, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25393 + }, + { + "epoch": 20.282747603833865, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 25394 + }, + { + "epoch": 20.283546325878593, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 25395 + }, + { + "epoch": 20.284345047923324, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 25396 + }, + { + "epoch": 20.28514376996805, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 25397 + }, + { + "epoch": 20.28594249201278, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 25398 + }, + { + "epoch": 20.286741214057507, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 25399 + }, + { + "epoch": 20.287539936102238, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 25400 + }, + { + "epoch": 20.288338658146966, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 25401 + }, + { + "epoch": 20.289137380191693, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 25402 + }, + { + "epoch": 20.28993610223642, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 25403 + }, + { + "epoch": 20.29073482428115, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 25404 + }, + { + "epoch": 20.29153354632588, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25405 + }, + { + "epoch": 20.292332268370608, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 25406 + }, + { + "epoch": 20.293130990415335, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 25407 + }, + { + "epoch": 20.293929712460063, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 25408 + }, + { + "epoch": 20.294728434504794, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 25409 + }, + { + "epoch": 20.29552715654952, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 25410 + }, + { + "epoch": 20.29632587859425, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 25411 + }, + { + "epoch": 20.297124600638977, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 25412 + }, + { + "epoch": 20.297923322683705, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 25413 + }, + { + "epoch": 20.298722044728436, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 25414 + }, + { + "epoch": 20.299520766773163, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 25415 + }, + { + "epoch": 20.30031948881789, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 25416 + }, + { + "epoch": 20.30111821086262, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25417 + }, + { + "epoch": 20.30191693290735, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 25418 + }, + { + "epoch": 20.302715654952078, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 25419 + }, + { + "epoch": 20.303514376996805, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 25420 + }, + { + "epoch": 20.304313099041533, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 25421 + }, + { + "epoch": 20.30511182108626, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 25422 + }, + { + "epoch": 20.30591054313099, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 25423 + }, + { + "epoch": 20.30670926517572, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 25424 + }, + { + "epoch": 20.307507987220447, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 25425 + }, + { + "epoch": 20.308306709265175, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 25426 + }, + { + "epoch": 20.309105431309906, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 25427 + }, + { + "epoch": 20.309904153354633, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 25428 + }, + { + "epoch": 20.31070287539936, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 25429 + }, + { + "epoch": 20.31150159744409, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 25430 + }, + { + "epoch": 20.312300319488816, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 25431 + }, + { + "epoch": 20.313099041533548, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 25432 + }, + { + "epoch": 20.313897763578275, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25433 + }, + { + "epoch": 20.314696485623003, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 25434 + }, + { + "epoch": 20.31549520766773, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 25435 + }, + { + "epoch": 20.31629392971246, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 25436 + }, + { + "epoch": 20.31709265175719, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 25437 + }, + { + "epoch": 20.317891373801917, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25438 + }, + { + "epoch": 20.318690095846645, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 25439 + }, + { + "epoch": 20.319488817891372, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 25440 + }, + { + "epoch": 20.320287539936103, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 25441 + }, + { + "epoch": 20.32108626198083, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 25442 + }, + { + "epoch": 20.32188498402556, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 25443 + }, + { + "epoch": 20.322683706070286, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25444 + }, + { + "epoch": 20.323482428115017, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 25445 + }, + { + "epoch": 20.324281150159745, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 25446 + }, + { + "epoch": 20.325079872204473, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 25447 + }, + { + "epoch": 20.3258785942492, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 25448 + }, + { + "epoch": 20.326677316293928, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25449 + }, + { + "epoch": 20.32747603833866, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 25450 + }, + { + "epoch": 20.328274760383387, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 25451 + }, + { + "epoch": 20.329073482428115, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 25452 + }, + { + "epoch": 20.329872204472842, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 25453 + }, + { + "epoch": 20.330670926517573, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 25454 + }, + { + "epoch": 20.3314696485623, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 25455 + }, + { + "epoch": 20.33226837060703, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 25456 + }, + { + "epoch": 20.333067092651756, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 25457 + }, + { + "epoch": 20.333865814696484, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 25458 + }, + { + "epoch": 20.334664536741215, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 25459 + }, + { + "epoch": 20.335463258785943, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 25460 + }, + { + "epoch": 20.33626198083067, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 25461 + }, + { + "epoch": 20.337060702875398, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 25462 + }, + { + "epoch": 20.33785942492013, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 25463 + }, + { + "epoch": 20.338658146964857, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 25464 + }, + { + "epoch": 20.339456869009584, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 25465 + }, + { + "epoch": 20.340255591054312, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 25466 + }, + { + "epoch": 20.341054313099043, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 25467 + }, + { + "epoch": 20.34185303514377, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 25468 + }, + { + "epoch": 20.3426517571885, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 25469 + }, + { + "epoch": 20.343450479233226, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 25470 + }, + { + "epoch": 20.344249201277954, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 25471 + }, + { + "epoch": 20.345047923322685, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 25472 + }, + { + "epoch": 20.345846645367413, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 25473 + }, + { + "epoch": 20.34664536741214, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 25474 + }, + { + "epoch": 20.347444089456868, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 25475 + }, + { + "epoch": 20.3482428115016, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 25476 + }, + { + "epoch": 20.349041533546327, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 25477 + }, + { + "epoch": 20.349840255591054, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 25478 + }, + { + "epoch": 20.350638977635782, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 25479 + }, + { + "epoch": 20.35143769968051, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 25480 + }, + { + "epoch": 20.35223642172524, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 25481 + }, + { + "epoch": 20.35303514376997, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 25482 + }, + { + "epoch": 20.353833865814696, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 25483 + }, + { + "epoch": 20.354632587859424, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 25484 + }, + { + "epoch": 20.355431309904155, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 25485 + }, + { + "epoch": 20.356230031948883, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 25486 + }, + { + "epoch": 20.35702875399361, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 25487 + }, + { + "epoch": 20.357827476038338, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 25488 + }, + { + "epoch": 20.358626198083066, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 25489 + }, + { + "epoch": 20.359424920127797, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 25490 + }, + { + "epoch": 20.360223642172524, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 25491 + }, + { + "epoch": 20.361022364217252, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 25492 + }, + { + "epoch": 20.36182108626198, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 25493 + }, + { + "epoch": 20.36261980830671, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1075, + "step": 25494 + }, + { + "epoch": 20.36341853035144, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 25495 + }, + { + "epoch": 20.364217252396166, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 25496 + }, + { + "epoch": 20.365015974440894, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 25497 + }, + { + "epoch": 20.36581469648562, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 25498 + }, + { + "epoch": 20.366613418530353, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 25499 + }, + { + "epoch": 20.36741214057508, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 25500 + }, + { + "epoch": 20.368210862619808, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 25501 + }, + { + "epoch": 20.369009584664536, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 25502 + }, + { + "epoch": 20.369808306709267, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25503 + }, + { + "epoch": 20.370607028753994, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 25504 + }, + { + "epoch": 20.371405750798722, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25505 + }, + { + "epoch": 20.37220447284345, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 25506 + }, + { + "epoch": 20.373003194888177, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25507 + }, + { + "epoch": 20.37380191693291, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 25508 + }, + { + "epoch": 20.374600638977636, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 25509 + }, + { + "epoch": 20.375399361022364, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 25510 + }, + { + "epoch": 20.37619808306709, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 25511 + }, + { + "epoch": 20.376996805111823, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 25512 + }, + { + "epoch": 20.37779552715655, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 25513 + }, + { + "epoch": 20.378594249201278, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 25514 + }, + { + "epoch": 20.379392971246006, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25515 + }, + { + "epoch": 20.380191693290733, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 25516 + }, + { + "epoch": 20.380990415335464, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 25517 + }, + { + "epoch": 20.381789137380192, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 25518 + }, + { + "epoch": 20.38258785942492, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 25519 + }, + { + "epoch": 20.383386581469647, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 25520 + }, + { + "epoch": 20.38418530351438, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 25521 + }, + { + "epoch": 20.384984025559106, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 25522 + }, + { + "epoch": 20.385782747603834, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 25523 + }, + { + "epoch": 20.38658146964856, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 25524 + }, + { + "epoch": 20.38738019169329, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 25525 + }, + { + "epoch": 20.38817891373802, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 25526 + }, + { + "epoch": 20.388977635782748, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25527 + }, + { + "epoch": 20.389776357827476, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 25528 + }, + { + "epoch": 20.390575079872203, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 25529 + }, + { + "epoch": 20.391373801916934, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 25530 + }, + { + "epoch": 20.392172523961662, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 25531 + }, + { + "epoch": 20.39297124600639, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 25532 + }, + { + "epoch": 20.393769968051117, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 25533 + }, + { + "epoch": 20.394568690095845, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25534 + }, + { + "epoch": 20.395367412140576, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 25535 + }, + { + "epoch": 20.396166134185304, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 25536 + }, + { + "epoch": 20.39696485623003, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25537 + }, + { + "epoch": 20.39776357827476, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 25538 + }, + { + "epoch": 20.39856230031949, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 25539 + }, + { + "epoch": 20.399361022364218, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 25540 + }, + { + "epoch": 20.400159744408946, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 25541 + }, + { + "epoch": 20.400958466453673, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 25542 + }, + { + "epoch": 20.4017571884984, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 25543 + }, + { + "epoch": 20.402555910543132, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25544 + }, + { + "epoch": 20.40335463258786, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25545 + }, + { + "epoch": 20.404153354632587, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 25546 + }, + { + "epoch": 20.404952076677315, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 25547 + }, + { + "epoch": 20.405750798722046, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 25548 + }, + { + "epoch": 20.406549520766774, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 25549 + }, + { + "epoch": 20.4073482428115, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 25550 + }, + { + "epoch": 20.40814696485623, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 25551 + }, + { + "epoch": 20.408945686900957, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 25552 + }, + { + "epoch": 20.409744408945688, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 25553 + }, + { + "epoch": 20.410543130990416, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 25554 + }, + { + "epoch": 20.411341853035143, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 25555 + }, + { + "epoch": 20.41214057507987, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25556 + }, + { + "epoch": 20.412939297124602, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 25557 + }, + { + "epoch": 20.41373801916933, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 25558 + }, + { + "epoch": 20.414536741214057, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 25559 + }, + { + "epoch": 20.415335463258785, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 25560 + }, + { + "epoch": 20.416134185303516, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 25561 + }, + { + "epoch": 20.416932907348244, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 25562 + }, + { + "epoch": 20.41773162939297, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 25563 + }, + { + "epoch": 20.4185303514377, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 25564 + }, + { + "epoch": 20.419329073482427, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 25565 + }, + { + "epoch": 20.420127795527158, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 25566 + }, + { + "epoch": 20.420926517571885, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 25567 + }, + { + "epoch": 20.421725239616613, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 25568 + }, + { + "epoch": 20.42252396166134, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25569 + }, + { + "epoch": 20.423322683706072, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 25570 + }, + { + "epoch": 20.4241214057508, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 25571 + }, + { + "epoch": 20.424920127795527, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 25572 + }, + { + "epoch": 20.425718849840255, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 25573 + }, + { + "epoch": 20.426517571884983, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 25574 + }, + { + "epoch": 20.427316293929714, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 25575 + }, + { + "epoch": 20.42811501597444, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 25576 + }, + { + "epoch": 20.42891373801917, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 25577 + }, + { + "epoch": 20.429712460063897, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 25578 + }, + { + "epoch": 20.430511182108628, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 25579 + }, + { + "epoch": 20.431309904153355, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 25580 + }, + { + "epoch": 20.432108626198083, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 25581 + }, + { + "epoch": 20.43290734824281, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 25582 + }, + { + "epoch": 20.43370607028754, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 25583 + }, + { + "epoch": 20.43450479233227, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 25584 + }, + { + "epoch": 20.435303514376997, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 25585 + }, + { + "epoch": 20.436102236421725, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 25586 + }, + { + "epoch": 20.436900958466452, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 25587 + }, + { + "epoch": 20.437699680511184, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 25588 + }, + { + "epoch": 20.43849840255591, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 25589 + }, + { + "epoch": 20.43929712460064, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 25590 + }, + { + "epoch": 20.440095846645367, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 25591 + }, + { + "epoch": 20.440894568690094, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 25592 + }, + { + "epoch": 20.441693290734825, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 25593 + }, + { + "epoch": 20.442492012779553, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 25594 + }, + { + "epoch": 20.44329073482428, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 25595 + }, + { + "epoch": 20.44408945686901, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 25596 + }, + { + "epoch": 20.44488817891374, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 25597 + }, + { + "epoch": 20.445686900958467, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 25598 + }, + { + "epoch": 20.446485623003195, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 25599 + }, + { + "epoch": 20.447284345047922, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 25600 + }, + { + "epoch": 20.44808306709265, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 25601 + }, + { + "epoch": 20.44888178913738, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 25602 + }, + { + "epoch": 20.44968051118211, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 25603 + }, + { + "epoch": 20.450479233226837, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25604 + }, + { + "epoch": 20.451277955271564, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 25605 + }, + { + "epoch": 20.452076677316295, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 25606 + }, + { + "epoch": 20.452875399361023, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25607 + }, + { + "epoch": 20.45367412140575, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 25608 + }, + { + "epoch": 20.45447284345048, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 25609 + }, + { + "epoch": 20.455271565495206, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 25610 + }, + { + "epoch": 20.456070287539937, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25611 + }, + { + "epoch": 20.456869009584665, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 25612 + }, + { + "epoch": 20.457667731629392, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 25613 + }, + { + "epoch": 20.45846645367412, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 25614 + }, + { + "epoch": 20.45926517571885, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 25615 + }, + { + "epoch": 20.46006389776358, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 25616 + }, + { + "epoch": 20.460862619808307, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 25617 + }, + { + "epoch": 20.461661341853034, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 25618 + }, + { + "epoch": 20.462460063897762, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 25619 + }, + { + "epoch": 20.463258785942493, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 25620 + }, + { + "epoch": 20.46405750798722, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 25621 + }, + { + "epoch": 20.46485623003195, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 25622 + }, + { + "epoch": 20.465654952076676, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 25623 + }, + { + "epoch": 20.466453674121407, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 25624 + }, + { + "epoch": 20.467252396166135, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25625 + }, + { + "epoch": 20.468051118210862, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 25626 + }, + { + "epoch": 20.46884984025559, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 25627 + }, + { + "epoch": 20.46964856230032, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 25628 + }, + { + "epoch": 20.47044728434505, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 25629 + }, + { + "epoch": 20.471246006389777, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 25630 + }, + { + "epoch": 20.472044728434504, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 25631 + }, + { + "epoch": 20.472843450479232, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 25632 + }, + { + "epoch": 20.473642172523963, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 25633 + }, + { + "epoch": 20.47444089456869, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 25634 + }, + { + "epoch": 20.47523961661342, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25635 + }, + { + "epoch": 20.476038338658146, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 25636 + }, + { + "epoch": 20.476837060702877, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 25637 + }, + { + "epoch": 20.477635782747605, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 25638 + }, + { + "epoch": 20.478434504792332, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 25639 + }, + { + "epoch": 20.47923322683706, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25640 + }, + { + "epoch": 20.480031948881788, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 25641 + }, + { + "epoch": 20.48083067092652, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 25642 + }, + { + "epoch": 20.481629392971247, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 25643 + }, + { + "epoch": 20.482428115015974, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 25644 + }, + { + "epoch": 20.483226837060702, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 25645 + }, + { + "epoch": 20.484025559105433, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 25646 + }, + { + "epoch": 20.48482428115016, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 25647 + }, + { + "epoch": 20.48562300319489, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 25648 + }, + { + "epoch": 20.486421725239616, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 25649 + }, + { + "epoch": 20.487220447284344, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25650 + }, + { + "epoch": 20.488019169329075, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0754, + "step": 25651 + }, + { + "epoch": 20.488817891373802, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 25652 + }, + { + "epoch": 20.48961661341853, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 25653 + }, + { + "epoch": 20.490415335463258, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 25654 + }, + { + "epoch": 20.49121405750799, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 25655 + }, + { + "epoch": 20.492012779552716, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 25656 + }, + { + "epoch": 20.492811501597444, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 25657 + }, + { + "epoch": 20.49361022364217, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 25658 + }, + { + "epoch": 20.4944089456869, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 25659 + }, + { + "epoch": 20.49520766773163, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 25660 + }, + { + "epoch": 20.49600638977636, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 25661 + }, + { + "epoch": 20.496805111821086, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 25662 + }, + { + "epoch": 20.497603833865814, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 25663 + }, + { + "epoch": 20.498402555910545, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 25664 + }, + { + "epoch": 20.499201277955272, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 25665 + }, + { + "epoch": 20.5, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 25666 + }, + { + "epoch": 20.500798722044728, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 25667 + }, + { + "epoch": 20.501597444089455, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 25668 + }, + { + "epoch": 20.502396166134186, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25669 + }, + { + "epoch": 20.503194888178914, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 25670 + }, + { + "epoch": 20.50399361022364, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 25671 + }, + { + "epoch": 20.50479233226837, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 25672 + }, + { + "epoch": 20.5055910543131, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 25673 + }, + { + "epoch": 20.50638977635783, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 25674 + }, + { + "epoch": 20.507188498402556, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 25675 + }, + { + "epoch": 20.507987220447284, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 25676 + }, + { + "epoch": 20.50878594249201, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 25677 + }, + { + "epoch": 20.509584664536742, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 25678 + }, + { + "epoch": 20.51038338658147, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 25679 + }, + { + "epoch": 20.511182108626198, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25680 + }, + { + "epoch": 20.511980830670925, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 25681 + }, + { + "epoch": 20.512779552715656, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 25682 + }, + { + "epoch": 20.513578274760384, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 25683 + }, + { + "epoch": 20.51437699680511, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 25684 + }, + { + "epoch": 20.51517571884984, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 25685 + }, + { + "epoch": 20.515974440894567, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 25686 + }, + { + "epoch": 20.516773162939298, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 25687 + }, + { + "epoch": 20.517571884984026, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 25688 + }, + { + "epoch": 20.518370607028753, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 25689 + }, + { + "epoch": 20.51916932907348, + "grad_norm": 0.341796875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 25690 + }, + { + "epoch": 20.519968051118212, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 25691 + }, + { + "epoch": 20.52076677316294, + "grad_norm": 0.318359375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 25692 + }, + { + "epoch": 20.521565495207668, + "grad_norm": 0.291015625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 25693 + }, + { + "epoch": 20.522364217252395, + "grad_norm": 0.357421875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 25694 + }, + { + "epoch": 20.523162939297123, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 25695 + }, + { + "epoch": 20.523961661341854, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 25696 + }, + { + "epoch": 20.52476038338658, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 25697 + }, + { + "epoch": 20.52555910543131, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 25698 + }, + { + "epoch": 20.526357827476037, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 25699 + }, + { + "epoch": 20.527156549520768, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 25700 + }, + { + "epoch": 20.527955271565496, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 25701 + }, + { + "epoch": 20.528753993610223, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 25702 + }, + { + "epoch": 20.52955271565495, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 25703 + }, + { + "epoch": 20.53035143769968, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25704 + }, + { + "epoch": 20.53115015974441, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 25705 + }, + { + "epoch": 20.531948881789138, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25706 + }, + { + "epoch": 20.532747603833865, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 25707 + }, + { + "epoch": 20.533546325878593, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 25708 + }, + { + "epoch": 20.534345047923324, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 25709 + }, + { + "epoch": 20.53514376996805, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 25710 + }, + { + "epoch": 20.53594249201278, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 25711 + }, + { + "epoch": 20.536741214057507, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 25712 + }, + { + "epoch": 20.537539936102235, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 25713 + }, + { + "epoch": 20.538338658146966, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 25714 + }, + { + "epoch": 20.539137380191693, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 25715 + }, + { + "epoch": 20.53993610223642, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 25716 + }, + { + "epoch": 20.54073482428115, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 25717 + }, + { + "epoch": 20.54153354632588, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 25718 + }, + { + "epoch": 20.542332268370608, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 25719 + }, + { + "epoch": 20.543130990415335, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 25720 + }, + { + "epoch": 20.543929712460063, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 25721 + }, + { + "epoch": 20.544728434504794, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 25722 + }, + { + "epoch": 20.54552715654952, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 25723 + }, + { + "epoch": 20.54632587859425, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25724 + }, + { + "epoch": 20.547124600638977, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 25725 + }, + { + "epoch": 20.547923322683705, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 25726 + }, + { + "epoch": 20.548722044728436, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 25727 + }, + { + "epoch": 20.549520766773163, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25728 + }, + { + "epoch": 20.55031948881789, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 25729 + }, + { + "epoch": 20.55111821086262, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25730 + }, + { + "epoch": 20.55191693290735, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 25731 + }, + { + "epoch": 20.552715654952078, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 25732 + }, + { + "epoch": 20.553514376996805, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 25733 + }, + { + "epoch": 20.554313099041533, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 25734 + }, + { + "epoch": 20.55511182108626, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 25735 + }, + { + "epoch": 20.55591054313099, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 25736 + }, + { + "epoch": 20.55670926517572, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 25737 + }, + { + "epoch": 20.557507987220447, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 25738 + }, + { + "epoch": 20.558306709265175, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 25739 + }, + { + "epoch": 20.559105431309906, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 25740 + }, + { + "epoch": 20.559904153354633, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 25741 + }, + { + "epoch": 20.56070287539936, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 25742 + }, + { + "epoch": 20.56150159744409, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 25743 + }, + { + "epoch": 20.562300319488816, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25744 + }, + { + "epoch": 20.563099041533548, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 25745 + }, + { + "epoch": 20.563897763578275, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 25746 + }, + { + "epoch": 20.564696485623003, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 25747 + }, + { + "epoch": 20.56549520766773, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 25748 + }, + { + "epoch": 20.56629392971246, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 25749 + }, + { + "epoch": 20.56709265175719, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 25750 + }, + { + "epoch": 20.567891373801917, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 25751 + }, + { + "epoch": 20.568690095846645, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 25752 + }, + { + "epoch": 20.569488817891372, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 25753 + }, + { + "epoch": 20.570287539936103, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 25754 + }, + { + "epoch": 20.57108626198083, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 25755 + }, + { + "epoch": 20.57188498402556, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 25756 + }, + { + "epoch": 20.572683706070286, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 25757 + }, + { + "epoch": 20.573482428115017, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25758 + }, + { + "epoch": 20.574281150159745, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 25759 + }, + { + "epoch": 20.575079872204473, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 25760 + }, + { + "epoch": 20.5758785942492, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 25761 + }, + { + "epoch": 20.576677316293928, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 25762 + }, + { + "epoch": 20.57747603833866, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 25763 + }, + { + "epoch": 20.578274760383387, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 25764 + }, + { + "epoch": 20.579073482428115, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 25765 + }, + { + "epoch": 20.579872204472842, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 25766 + }, + { + "epoch": 20.580670926517573, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 25767 + }, + { + "epoch": 20.5814696485623, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 25768 + }, + { + "epoch": 20.58226837060703, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 25769 + }, + { + "epoch": 20.583067092651756, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 25770 + }, + { + "epoch": 20.583865814696484, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 25771 + }, + { + "epoch": 20.584664536741215, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25772 + }, + { + "epoch": 20.585463258785943, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 25773 + }, + { + "epoch": 20.58626198083067, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 25774 + }, + { + "epoch": 20.587060702875398, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 25775 + }, + { + "epoch": 20.58785942492013, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 25776 + }, + { + "epoch": 20.588658146964857, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 25777 + }, + { + "epoch": 20.589456869009584, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 25778 + }, + { + "epoch": 20.590255591054312, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 25779 + }, + { + "epoch": 20.591054313099043, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 25780 + }, + { + "epoch": 20.59185303514377, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 25781 + }, + { + "epoch": 20.5926517571885, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25782 + }, + { + "epoch": 20.593450479233226, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 25783 + }, + { + "epoch": 20.594249201277954, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 25784 + }, + { + "epoch": 20.595047923322685, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 25785 + }, + { + "epoch": 20.595846645367413, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 25786 + }, + { + "epoch": 20.59664536741214, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25787 + }, + { + "epoch": 20.597444089456868, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 25788 + }, + { + "epoch": 20.5982428115016, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 25789 + }, + { + "epoch": 20.599041533546327, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 25790 + }, + { + "epoch": 20.599840255591054, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 25791 + }, + { + "epoch": 20.600638977635782, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 25792 + }, + { + "epoch": 20.60143769968051, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0762, + "step": 25793 + }, + { + "epoch": 20.60223642172524, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 25794 + }, + { + "epoch": 20.60303514376997, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 25795 + }, + { + "epoch": 20.603833865814696, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25796 + }, + { + "epoch": 20.604632587859424, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 25797 + }, + { + "epoch": 20.605431309904155, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 25798 + }, + { + "epoch": 20.606230031948883, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25799 + }, + { + "epoch": 20.60702875399361, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 25800 + }, + { + "epoch": 20.607827476038338, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 25801 + }, + { + "epoch": 20.608626198083066, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 25802 + }, + { + "epoch": 20.609424920127797, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25803 + }, + { + "epoch": 20.610223642172524, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 25804 + }, + { + "epoch": 20.611022364217252, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25805 + }, + { + "epoch": 20.61182108626198, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 25806 + }, + { + "epoch": 20.61261980830671, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 25807 + }, + { + "epoch": 20.61341853035144, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0754, + "step": 25808 + }, + { + "epoch": 20.614217252396166, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 25809 + }, + { + "epoch": 20.615015974440894, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 25810 + }, + { + "epoch": 20.61581469648562, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 25811 + }, + { + "epoch": 20.616613418530353, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 25812 + }, + { + "epoch": 20.61741214057508, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 25813 + }, + { + "epoch": 20.618210862619808, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 25814 + }, + { + "epoch": 20.619009584664536, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25815 + }, + { + "epoch": 20.619808306709267, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 25816 + }, + { + "epoch": 20.620607028753994, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 25817 + }, + { + "epoch": 20.621405750798722, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 25818 + }, + { + "epoch": 20.62220447284345, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 25819 + }, + { + "epoch": 20.623003194888177, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 25820 + }, + { + "epoch": 20.62380191693291, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 25821 + }, + { + "epoch": 20.624600638977636, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 25822 + }, + { + "epoch": 20.625399361022364, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 25823 + }, + { + "epoch": 20.62619808306709, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 25824 + }, + { + "epoch": 20.626996805111823, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 25825 + }, + { + "epoch": 20.62779552715655, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 25826 + }, + { + "epoch": 20.628594249201278, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 25827 + }, + { + "epoch": 20.629392971246006, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 25828 + }, + { + "epoch": 20.630191693290733, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 25829 + }, + { + "epoch": 20.630990415335464, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 25830 + }, + { + "epoch": 20.631789137380192, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 25831 + }, + { + "epoch": 20.63258785942492, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 25832 + }, + { + "epoch": 20.633386581469647, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 25833 + }, + { + "epoch": 20.63418530351438, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 25834 + }, + { + "epoch": 20.634984025559106, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 25835 + }, + { + "epoch": 20.635782747603834, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 25836 + }, + { + "epoch": 20.63658146964856, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 25837 + }, + { + "epoch": 20.63738019169329, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 25838 + }, + { + "epoch": 20.63817891373802, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 25839 + }, + { + "epoch": 20.638977635782748, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25840 + }, + { + "epoch": 20.639776357827476, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 25841 + }, + { + "epoch": 20.640575079872203, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 25842 + }, + { + "epoch": 20.641373801916934, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 25843 + }, + { + "epoch": 20.642172523961662, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 25844 + }, + { + "epoch": 20.64297124600639, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 25845 + }, + { + "epoch": 20.643769968051117, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 25846 + }, + { + "epoch": 20.644568690095845, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 25847 + }, + { + "epoch": 20.645367412140576, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 25848 + }, + { + "epoch": 20.646166134185304, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 25849 + }, + { + "epoch": 20.64696485623003, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 25850 + }, + { + "epoch": 20.64776357827476, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25851 + }, + { + "epoch": 20.64856230031949, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 25852 + }, + { + "epoch": 20.649361022364218, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25853 + }, + { + "epoch": 20.650159744408946, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 25854 + }, + { + "epoch": 20.650958466453673, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 25855 + }, + { + "epoch": 20.6517571884984, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 25856 + }, + { + "epoch": 20.652555910543132, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 25857 + }, + { + "epoch": 20.65335463258786, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 25858 + }, + { + "epoch": 20.654153354632587, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 25859 + }, + { + "epoch": 20.654952076677315, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 25860 + }, + { + "epoch": 20.655750798722046, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 25861 + }, + { + "epoch": 20.656549520766774, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 25862 + }, + { + "epoch": 20.6573482428115, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 25863 + }, + { + "epoch": 20.65814696485623, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 25864 + }, + { + "epoch": 20.658945686900957, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 25865 + }, + { + "epoch": 20.659744408945688, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 25866 + }, + { + "epoch": 20.660543130990416, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 25867 + }, + { + "epoch": 20.661341853035143, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 25868 + }, + { + "epoch": 20.66214057507987, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 25869 + }, + { + "epoch": 20.662939297124602, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 25870 + }, + { + "epoch": 20.66373801916933, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 25871 + }, + { + "epoch": 20.664536741214057, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25872 + }, + { + "epoch": 20.665335463258785, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 25873 + }, + { + "epoch": 20.666134185303516, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 25874 + }, + { + "epoch": 20.666932907348244, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 25875 + }, + { + "epoch": 20.66773162939297, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 25876 + }, + { + "epoch": 20.6685303514377, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 25877 + }, + { + "epoch": 20.669329073482427, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 25878 + }, + { + "epoch": 20.670127795527158, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 25879 + }, + { + "epoch": 20.670926517571885, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 25880 + }, + { + "epoch": 20.671725239616613, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 25881 + }, + { + "epoch": 20.67252396166134, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 25882 + }, + { + "epoch": 20.673322683706072, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 25883 + }, + { + "epoch": 20.6741214057508, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 25884 + }, + { + "epoch": 20.674920127795527, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 25885 + }, + { + "epoch": 20.675718849840255, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 25886 + }, + { + "epoch": 20.676517571884983, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 25887 + }, + { + "epoch": 20.677316293929714, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 25888 + }, + { + "epoch": 20.67811501597444, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 25889 + }, + { + "epoch": 20.67891373801917, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 25890 + }, + { + "epoch": 20.679712460063897, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 25891 + }, + { + "epoch": 20.680511182108628, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25892 + }, + { + "epoch": 20.681309904153355, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 25893 + }, + { + "epoch": 20.682108626198083, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 25894 + }, + { + "epoch": 20.68290734824281, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 25895 + }, + { + "epoch": 20.68370607028754, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 25896 + }, + { + "epoch": 20.68450479233227, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 25897 + }, + { + "epoch": 20.685303514376997, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 25898 + }, + { + "epoch": 20.686102236421725, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 25899 + }, + { + "epoch": 20.686900958466452, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 25900 + }, + { + "epoch": 20.687699680511184, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25901 + }, + { + "epoch": 20.68849840255591, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 25902 + }, + { + "epoch": 20.68929712460064, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 25903 + }, + { + "epoch": 20.690095846645367, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 25904 + }, + { + "epoch": 20.690894568690094, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 25905 + }, + { + "epoch": 20.691693290734825, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25906 + }, + { + "epoch": 20.692492012779553, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 25907 + }, + { + "epoch": 20.69329073482428, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 25908 + }, + { + "epoch": 20.69408945686901, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 25909 + }, + { + "epoch": 20.69488817891374, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 25910 + }, + { + "epoch": 20.695686900958467, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 25911 + }, + { + "epoch": 20.696485623003195, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25912 + }, + { + "epoch": 20.697284345047922, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 25913 + }, + { + "epoch": 20.69808306709265, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 25914 + }, + { + "epoch": 20.69888178913738, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 25915 + }, + { + "epoch": 20.69968051118211, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 25916 + }, + { + "epoch": 20.700479233226837, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 25917 + }, + { + "epoch": 20.701277955271564, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 25918 + }, + { + "epoch": 20.702076677316295, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 25919 + }, + { + "epoch": 20.702875399361023, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 25920 + }, + { + "epoch": 20.70367412140575, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 25921 + }, + { + "epoch": 20.70447284345048, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 25922 + }, + { + "epoch": 20.705271565495206, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 25923 + }, + { + "epoch": 20.706070287539937, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 25924 + }, + { + "epoch": 20.706869009584665, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 25925 + }, + { + "epoch": 20.707667731629392, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 25926 + }, + { + "epoch": 20.70846645367412, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 25927 + }, + { + "epoch": 20.70926517571885, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 25928 + }, + { + "epoch": 20.71006389776358, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 25929 + }, + { + "epoch": 20.710862619808307, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25930 + }, + { + "epoch": 20.711661341853034, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 25931 + }, + { + "epoch": 20.712460063897765, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 25932 + }, + { + "epoch": 20.713258785942493, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 25933 + }, + { + "epoch": 20.71405750798722, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 25934 + }, + { + "epoch": 20.71485623003195, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 25935 + }, + { + "epoch": 20.715654952076676, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 25936 + }, + { + "epoch": 20.716453674121407, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 25937 + }, + { + "epoch": 20.717252396166135, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 25938 + }, + { + "epoch": 20.718051118210862, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 25939 + }, + { + "epoch": 20.71884984025559, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25940 + }, + { + "epoch": 20.71964856230032, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 25941 + }, + { + "epoch": 20.72044728434505, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 25942 + }, + { + "epoch": 20.721246006389777, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 25943 + }, + { + "epoch": 20.722044728434504, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 25944 + }, + { + "epoch": 20.722843450479232, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 25945 + }, + { + "epoch": 20.723642172523963, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 25946 + }, + { + "epoch": 20.72444089456869, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 25947 + }, + { + "epoch": 20.72523961661342, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 25948 + }, + { + "epoch": 20.726038338658146, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 25949 + }, + { + "epoch": 20.726837060702877, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 25950 + }, + { + "epoch": 20.727635782747605, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 25951 + }, + { + "epoch": 20.728434504792332, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 25952 + }, + { + "epoch": 20.72923322683706, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 25953 + }, + { + "epoch": 20.730031948881788, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 25954 + }, + { + "epoch": 20.73083067092652, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 25955 + }, + { + "epoch": 20.731629392971247, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 25956 + }, + { + "epoch": 20.732428115015974, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 25957 + }, + { + "epoch": 20.733226837060702, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 25958 + }, + { + "epoch": 20.734025559105433, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 25959 + }, + { + "epoch": 20.73482428115016, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 25960 + }, + { + "epoch": 20.73562300319489, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 25961 + }, + { + "epoch": 20.736421725239616, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 25962 + }, + { + "epoch": 20.737220447284344, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 25963 + }, + { + "epoch": 20.738019169329075, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 25964 + }, + { + "epoch": 20.738817891373802, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 25965 + }, + { + "epoch": 20.73961661341853, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 25966 + }, + { + "epoch": 20.740415335463258, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 25967 + }, + { + "epoch": 20.74121405750799, + "grad_norm": 0.32421875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 25968 + }, + { + "epoch": 20.742012779552716, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 25969 + }, + { + "epoch": 20.742811501597444, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 25970 + }, + { + "epoch": 20.74361022364217, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 25971 + }, + { + "epoch": 20.7444089456869, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 25972 + }, + { + "epoch": 20.74520766773163, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 25973 + }, + { + "epoch": 20.74600638977636, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 25974 + }, + { + "epoch": 20.746805111821086, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 25975 + }, + { + "epoch": 20.747603833865814, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 25976 + }, + { + "epoch": 20.748402555910545, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 25977 + }, + { + "epoch": 20.749201277955272, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 25978 + }, + { + "epoch": 20.75, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 25979 + }, + { + "epoch": 20.750798722044728, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 25980 + }, + { + "epoch": 20.751597444089455, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 25981 + }, + { + "epoch": 20.752396166134186, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 25982 + }, + { + "epoch": 20.753194888178914, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 25983 + }, + { + "epoch": 20.75399361022364, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 25984 + }, + { + "epoch": 20.75479233226837, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 25985 + }, + { + "epoch": 20.7555910543131, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 25986 + }, + { + "epoch": 20.75638977635783, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 25987 + }, + { + "epoch": 20.757188498402556, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 25988 + }, + { + "epoch": 20.757987220447284, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 25989 + }, + { + "epoch": 20.75878594249201, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 25990 + }, + { + "epoch": 20.759584664536742, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 25991 + }, + { + "epoch": 20.76038338658147, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 25992 + }, + { + "epoch": 20.761182108626198, + "grad_norm": 0.376953125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 25993 + }, + { + "epoch": 20.761980830670925, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 25994 + }, + { + "epoch": 20.762779552715656, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 25995 + }, + { + "epoch": 20.763578274760384, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 25996 + }, + { + "epoch": 20.76437699680511, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 25997 + }, + { + "epoch": 20.76517571884984, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 25998 + }, + { + "epoch": 20.765974440894567, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 25999 + }, + { + "epoch": 20.766773162939298, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 26000 + }, + { + "epoch": 20.767571884984026, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 26001 + }, + { + "epoch": 20.768370607028753, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 26002 + }, + { + "epoch": 20.76916932907348, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 26003 + }, + { + "epoch": 20.769968051118212, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 26004 + }, + { + "epoch": 20.77076677316294, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 26005 + }, + { + "epoch": 20.771565495207668, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 26006 + }, + { + "epoch": 20.772364217252395, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 26007 + }, + { + "epoch": 20.773162939297123, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 26008 + }, + { + "epoch": 20.773961661341854, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 26009 + }, + { + "epoch": 20.77476038338658, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26010 + }, + { + "epoch": 20.77555910543131, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 26011 + }, + { + "epoch": 20.776357827476037, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 26012 + }, + { + "epoch": 20.777156549520768, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 26013 + }, + { + "epoch": 20.777955271565496, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26014 + }, + { + "epoch": 20.778753993610223, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 26015 + }, + { + "epoch": 20.77955271565495, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 26016 + }, + { + "epoch": 20.78035143769968, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 26017 + }, + { + "epoch": 20.78115015974441, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 26018 + }, + { + "epoch": 20.781948881789138, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 26019 + }, + { + "epoch": 20.782747603833865, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 26020 + }, + { + "epoch": 20.783546325878593, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 26021 + }, + { + "epoch": 20.784345047923324, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26022 + }, + { + "epoch": 20.78514376996805, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26023 + }, + { + "epoch": 20.78594249201278, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 26024 + }, + { + "epoch": 20.786741214057507, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 26025 + }, + { + "epoch": 20.787539936102235, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 26026 + }, + { + "epoch": 20.788338658146966, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 26027 + }, + { + "epoch": 20.789137380191693, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 26028 + }, + { + "epoch": 20.78993610223642, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 26029 + }, + { + "epoch": 20.79073482428115, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 26030 + }, + { + "epoch": 20.79153354632588, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 26031 + }, + { + "epoch": 20.792332268370608, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26032 + }, + { + "epoch": 20.793130990415335, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 26033 + }, + { + "epoch": 20.793929712460063, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 26034 + }, + { + "epoch": 20.794728434504794, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 26035 + }, + { + "epoch": 20.79552715654952, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 26036 + }, + { + "epoch": 20.79632587859425, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 26037 + }, + { + "epoch": 20.797124600638977, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 26038 + }, + { + "epoch": 20.797923322683705, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 26039 + }, + { + "epoch": 20.798722044728436, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 26040 + }, + { + "epoch": 20.799520766773163, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 26041 + }, + { + "epoch": 20.80031948881789, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 26042 + }, + { + "epoch": 20.80111821086262, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 26043 + }, + { + "epoch": 20.80191693290735, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26044 + }, + { + "epoch": 20.802715654952078, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 26045 + }, + { + "epoch": 20.803514376996805, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 26046 + }, + { + "epoch": 20.804313099041533, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 26047 + }, + { + "epoch": 20.80511182108626, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 26048 + }, + { + "epoch": 20.80591054313099, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26049 + }, + { + "epoch": 20.80670926517572, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1082, + "step": 26050 + }, + { + "epoch": 20.807507987220447, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 26051 + }, + { + "epoch": 20.808306709265175, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26052 + }, + { + "epoch": 20.809105431309906, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 26053 + }, + { + "epoch": 20.809904153354633, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 26054 + }, + { + "epoch": 20.81070287539936, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 26055 + }, + { + "epoch": 20.81150159744409, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1054, + "step": 26056 + }, + { + "epoch": 20.812300319488816, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 26057 + }, + { + "epoch": 20.813099041533548, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 26058 + }, + { + "epoch": 20.813897763578275, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 26059 + }, + { + "epoch": 20.814696485623003, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 26060 + }, + { + "epoch": 20.81549520766773, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 26061 + }, + { + "epoch": 20.81629392971246, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 26062 + }, + { + "epoch": 20.81709265175719, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 26063 + }, + { + "epoch": 20.817891373801917, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 26064 + }, + { + "epoch": 20.818690095846645, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 26065 + }, + { + "epoch": 20.819488817891372, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 26066 + }, + { + "epoch": 20.820287539936103, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 26067 + }, + { + "epoch": 20.82108626198083, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 26068 + }, + { + "epoch": 20.82188498402556, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 26069 + }, + { + "epoch": 20.822683706070286, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 26070 + }, + { + "epoch": 20.823482428115017, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26071 + }, + { + "epoch": 20.824281150159745, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 26072 + }, + { + "epoch": 20.825079872204473, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 26073 + }, + { + "epoch": 20.8258785942492, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 26074 + }, + { + "epoch": 20.826677316293928, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 26075 + }, + { + "epoch": 20.82747603833866, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 26076 + }, + { + "epoch": 20.828274760383387, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 26077 + }, + { + "epoch": 20.829073482428115, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 26078 + }, + { + "epoch": 20.829872204472842, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 26079 + }, + { + "epoch": 20.830670926517573, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 26080 + }, + { + "epoch": 20.8314696485623, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 26081 + }, + { + "epoch": 20.83226837060703, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 26082 + }, + { + "epoch": 20.833067092651756, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 26083 + }, + { + "epoch": 20.833865814696484, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 26084 + }, + { + "epoch": 20.834664536741215, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 26085 + }, + { + "epoch": 20.835463258785943, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 26086 + }, + { + "epoch": 20.83626198083067, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 26087 + }, + { + "epoch": 20.837060702875398, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 26088 + }, + { + "epoch": 20.83785942492013, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 26089 + }, + { + "epoch": 20.838658146964857, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1068, + "step": 26090 + }, + { + "epoch": 20.839456869009584, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 26091 + }, + { + "epoch": 20.840255591054312, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 26092 + }, + { + "epoch": 20.841054313099043, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 26093 + }, + { + "epoch": 20.84185303514377, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26094 + }, + { + "epoch": 20.8426517571885, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26095 + }, + { + "epoch": 20.843450479233226, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 26096 + }, + { + "epoch": 20.844249201277954, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 26097 + }, + { + "epoch": 20.845047923322685, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 26098 + }, + { + "epoch": 20.845846645367413, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26099 + }, + { + "epoch": 20.84664536741214, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 26100 + }, + { + "epoch": 20.847444089456868, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 26101 + }, + { + "epoch": 20.8482428115016, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 26102 + }, + { + "epoch": 20.849041533546327, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 26103 + }, + { + "epoch": 20.849840255591054, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 26104 + }, + { + "epoch": 20.850638977635782, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 26105 + }, + { + "epoch": 20.85143769968051, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 26106 + }, + { + "epoch": 20.85223642172524, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 26107 + }, + { + "epoch": 20.85303514376997, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26108 + }, + { + "epoch": 20.853833865814696, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 26109 + }, + { + "epoch": 20.854632587859424, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 26110 + }, + { + "epoch": 20.855431309904155, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 26111 + }, + { + "epoch": 20.856230031948883, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 26112 + }, + { + "epoch": 20.85702875399361, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 26113 + }, + { + "epoch": 20.857827476038338, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.1042, + "step": 26114 + }, + { + "epoch": 20.858626198083066, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 26115 + }, + { + "epoch": 20.859424920127797, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 26116 + }, + { + "epoch": 20.860223642172524, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 26117 + }, + { + "epoch": 20.861022364217252, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 26118 + }, + { + "epoch": 20.86182108626198, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 26119 + }, + { + "epoch": 20.86261980830671, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 26120 + }, + { + "epoch": 20.86341853035144, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 26121 + }, + { + "epoch": 20.864217252396166, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 26122 + }, + { + "epoch": 20.865015974440894, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 26123 + }, + { + "epoch": 20.86581469648562, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 26124 + }, + { + "epoch": 20.866613418530353, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 26125 + }, + { + "epoch": 20.86741214057508, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 26126 + }, + { + "epoch": 20.868210862619808, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 26127 + }, + { + "epoch": 20.869009584664536, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 26128 + }, + { + "epoch": 20.869808306709267, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26129 + }, + { + "epoch": 20.870607028753994, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 26130 + }, + { + "epoch": 20.871405750798722, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 26131 + }, + { + "epoch": 20.87220447284345, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 26132 + }, + { + "epoch": 20.873003194888177, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 26133 + }, + { + "epoch": 20.87380191693291, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 26134 + }, + { + "epoch": 20.874600638977636, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 26135 + }, + { + "epoch": 20.875399361022364, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26136 + }, + { + "epoch": 20.87619808306709, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 26137 + }, + { + "epoch": 20.876996805111823, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 26138 + }, + { + "epoch": 20.87779552715655, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 26139 + }, + { + "epoch": 20.878594249201278, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 26140 + }, + { + "epoch": 20.879392971246006, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 26141 + }, + { + "epoch": 20.880191693290733, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 26142 + }, + { + "epoch": 20.880990415335464, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 26143 + }, + { + "epoch": 20.881789137380192, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 26144 + }, + { + "epoch": 20.88258785942492, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 26145 + }, + { + "epoch": 20.883386581469647, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 26146 + }, + { + "epoch": 20.88418530351438, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 26147 + }, + { + "epoch": 20.884984025559106, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26148 + }, + { + "epoch": 20.885782747603834, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 26149 + }, + { + "epoch": 20.88658146964856, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 26150 + }, + { + "epoch": 20.88738019169329, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26151 + }, + { + "epoch": 20.88817891373802, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 26152 + }, + { + "epoch": 20.888977635782748, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 26153 + }, + { + "epoch": 20.889776357827476, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 26154 + }, + { + "epoch": 20.890575079872203, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 26155 + }, + { + "epoch": 20.891373801916934, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 26156 + }, + { + "epoch": 20.892172523961662, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 26157 + }, + { + "epoch": 20.89297124600639, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 26158 + }, + { + "epoch": 20.893769968051117, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 26159 + }, + { + "epoch": 20.894568690095845, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 26160 + }, + { + "epoch": 20.895367412140576, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 26161 + }, + { + "epoch": 20.896166134185304, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 26162 + }, + { + "epoch": 20.89696485623003, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26163 + }, + { + "epoch": 20.89776357827476, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 26164 + }, + { + "epoch": 20.89856230031949, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 26165 + }, + { + "epoch": 20.899361022364218, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 26166 + }, + { + "epoch": 20.900159744408946, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 26167 + }, + { + "epoch": 20.900958466453673, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26168 + }, + { + "epoch": 20.9017571884984, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 26169 + }, + { + "epoch": 20.902555910543132, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 26170 + }, + { + "epoch": 20.90335463258786, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 26171 + }, + { + "epoch": 20.904153354632587, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 26172 + }, + { + "epoch": 20.904952076677315, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 26173 + }, + { + "epoch": 20.905750798722046, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 26174 + }, + { + "epoch": 20.906549520766774, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 26175 + }, + { + "epoch": 20.9073482428115, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 26176 + }, + { + "epoch": 20.90814696485623, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 26177 + }, + { + "epoch": 20.908945686900957, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 26178 + }, + { + "epoch": 20.909744408945688, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 26179 + }, + { + "epoch": 20.910543130990416, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 26180 + }, + { + "epoch": 20.911341853035143, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 26181 + }, + { + "epoch": 20.91214057507987, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 26182 + }, + { + "epoch": 20.912939297124602, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 26183 + }, + { + "epoch": 20.91373801916933, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 26184 + }, + { + "epoch": 20.914536741214057, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 26185 + }, + { + "epoch": 20.915335463258785, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 26186 + }, + { + "epoch": 20.916134185303516, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 26187 + }, + { + "epoch": 20.916932907348244, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 26188 + }, + { + "epoch": 20.91773162939297, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 26189 + }, + { + "epoch": 20.9185303514377, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 26190 + }, + { + "epoch": 20.919329073482427, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26191 + }, + { + "epoch": 20.920127795527158, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 26192 + }, + { + "epoch": 20.920926517571885, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 26193 + }, + { + "epoch": 20.921725239616613, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26194 + }, + { + "epoch": 20.92252396166134, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 26195 + }, + { + "epoch": 20.923322683706072, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 26196 + }, + { + "epoch": 20.9241214057508, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 26197 + }, + { + "epoch": 20.924920127795527, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 26198 + }, + { + "epoch": 20.925718849840255, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 26199 + }, + { + "epoch": 20.926517571884983, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 26200 + }, + { + "epoch": 20.927316293929714, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 26201 + }, + { + "epoch": 20.92811501597444, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 26202 + }, + { + "epoch": 20.92891373801917, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 26203 + }, + { + "epoch": 20.929712460063897, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 26204 + }, + { + "epoch": 20.930511182108628, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 26205 + }, + { + "epoch": 20.931309904153355, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 26206 + }, + { + "epoch": 20.932108626198083, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 26207 + }, + { + "epoch": 20.93290734824281, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 26208 + }, + { + "epoch": 20.93370607028754, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 26209 + }, + { + "epoch": 20.93450479233227, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 26210 + }, + { + "epoch": 20.935303514376997, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 26211 + }, + { + "epoch": 20.936102236421725, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 26212 + }, + { + "epoch": 20.936900958466452, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 26213 + }, + { + "epoch": 20.937699680511184, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 26214 + }, + { + "epoch": 20.93849840255591, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26215 + }, + { + "epoch": 20.93929712460064, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 26216 + }, + { + "epoch": 20.940095846645367, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 26217 + }, + { + "epoch": 20.940894568690094, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26218 + }, + { + "epoch": 20.941693290734825, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 26219 + }, + { + "epoch": 20.942492012779553, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 26220 + }, + { + "epoch": 20.94329073482428, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 26221 + }, + { + "epoch": 20.94408945686901, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 26222 + }, + { + "epoch": 20.94488817891374, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 26223 + }, + { + "epoch": 20.945686900958467, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 26224 + }, + { + "epoch": 20.946485623003195, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 26225 + }, + { + "epoch": 20.947284345047922, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 26226 + }, + { + "epoch": 20.94808306709265, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26227 + }, + { + "epoch": 20.94888178913738, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 26228 + }, + { + "epoch": 20.94968051118211, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 26229 + }, + { + "epoch": 20.950479233226837, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 26230 + }, + { + "epoch": 20.951277955271564, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 26231 + }, + { + "epoch": 20.952076677316295, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 26232 + }, + { + "epoch": 20.952875399361023, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 26233 + }, + { + "epoch": 20.95367412140575, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 26234 + }, + { + "epoch": 20.95447284345048, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26235 + }, + { + "epoch": 20.955271565495206, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 26236 + }, + { + "epoch": 20.956070287539937, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 26237 + }, + { + "epoch": 20.956869009584665, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 26238 + }, + { + "epoch": 20.957667731629392, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 26239 + }, + { + "epoch": 20.95846645367412, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 26240 + }, + { + "epoch": 20.95926517571885, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 26241 + }, + { + "epoch": 20.96006389776358, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 26242 + }, + { + "epoch": 20.960862619808307, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 26243 + }, + { + "epoch": 20.961661341853034, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 26244 + }, + { + "epoch": 20.962460063897765, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 26245 + }, + { + "epoch": 20.963258785942493, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 26246 + }, + { + "epoch": 20.96405750798722, + "grad_norm": 0.361328125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 26247 + }, + { + "epoch": 20.96485623003195, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 26248 + }, + { + "epoch": 20.965654952076676, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 26249 + }, + { + "epoch": 20.966453674121407, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 26250 + }, + { + "epoch": 20.967252396166135, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 26251 + }, + { + "epoch": 20.968051118210862, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26252 + }, + { + "epoch": 20.96884984025559, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 26253 + }, + { + "epoch": 20.96964856230032, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 26254 + }, + { + "epoch": 20.97044728434505, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 26255 + }, + { + "epoch": 20.971246006389777, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 26256 + }, + { + "epoch": 20.972044728434504, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 26257 + }, + { + "epoch": 20.972843450479232, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 26258 + }, + { + "epoch": 20.973642172523963, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26259 + }, + { + "epoch": 20.97444089456869, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 26260 + }, + { + "epoch": 20.97523961661342, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 26261 + }, + { + "epoch": 20.976038338658146, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 26262 + }, + { + "epoch": 20.976837060702877, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 26263 + }, + { + "epoch": 20.977635782747605, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 26264 + }, + { + "epoch": 20.978434504792332, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26265 + }, + { + "epoch": 20.97923322683706, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 26266 + }, + { + "epoch": 20.980031948881788, + "grad_norm": 0.283203125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 26267 + }, + { + "epoch": 20.98083067092652, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 26268 + }, + { + "epoch": 20.981629392971247, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 26269 + }, + { + "epoch": 20.982428115015974, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 26270 + }, + { + "epoch": 20.983226837060702, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 26271 + }, + { + "epoch": 20.984025559105433, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 26272 + }, + { + "epoch": 20.98482428115016, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 26273 + }, + { + "epoch": 20.98562300319489, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 26274 + }, + { + "epoch": 20.986421725239616, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 26275 + }, + { + "epoch": 20.987220447284344, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 26276 + }, + { + "epoch": 20.988019169329075, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 26277 + }, + { + "epoch": 20.988817891373802, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 26278 + }, + { + "epoch": 20.98961661341853, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 26279 + }, + { + "epoch": 20.990415335463258, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 26280 + }, + { + "epoch": 20.99121405750799, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 26281 + }, + { + "epoch": 20.992012779552716, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 26282 + }, + { + "epoch": 20.992811501597444, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 26283 + }, + { + "epoch": 20.99361022364217, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 26284 + }, + { + "epoch": 20.9944089456869, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 26285 + }, + { + "epoch": 20.99520766773163, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0756, + "step": 26286 + }, + { + "epoch": 20.99600638977636, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 26287 + }, + { + "epoch": 20.996805111821086, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 26288 + }, + { + "epoch": 20.997603833865814, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 26289 + }, + { + "epoch": 20.998402555910545, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 26290 + }, + { + "epoch": 20.999201277955272, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 26291 + }, + { + "epoch": 21.0, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26292 + }, + { + "epoch": 21.000798722044728, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 26293 + }, + { + "epoch": 21.001597444089455, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 26294 + }, + { + "epoch": 21.002396166134186, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 26295 + }, + { + "epoch": 21.003194888178914, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 26296 + }, + { + "epoch": 21.00399361022364, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 26297 + }, + { + "epoch": 21.00479233226837, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 26298 + }, + { + "epoch": 21.0055910543131, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 26299 + }, + { + "epoch": 21.00638977635783, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 26300 + }, + { + "epoch": 21.007188498402556, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 26301 + }, + { + "epoch": 21.007987220447284, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26302 + }, + { + "epoch": 21.00878594249201, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 26303 + }, + { + "epoch": 21.009584664536742, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26304 + }, + { + "epoch": 21.01038338658147, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 26305 + }, + { + "epoch": 21.011182108626198, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 26306 + }, + { + "epoch": 21.011980830670925, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26307 + }, + { + "epoch": 21.012779552715656, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 26308 + }, + { + "epoch": 21.013578274760384, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 26309 + }, + { + "epoch": 21.01437699680511, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 26310 + }, + { + "epoch": 21.01517571884984, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 26311 + }, + { + "epoch": 21.015974440894567, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26312 + }, + { + "epoch": 21.016773162939298, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26313 + }, + { + "epoch": 21.017571884984026, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 26314 + }, + { + "epoch": 21.018370607028753, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 26315 + }, + { + "epoch": 21.01916932907348, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 26316 + }, + { + "epoch": 21.019968051118212, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 26317 + }, + { + "epoch": 21.02076677316294, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 26318 + }, + { + "epoch": 21.021565495207668, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 26319 + }, + { + "epoch": 21.022364217252395, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 26320 + }, + { + "epoch": 21.023162939297123, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 26321 + }, + { + "epoch": 21.023961661341854, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 26322 + }, + { + "epoch": 21.02476038338658, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 26323 + }, + { + "epoch": 21.02555910543131, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 26324 + }, + { + "epoch": 21.026357827476037, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 26325 + }, + { + "epoch": 21.027156549520768, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26326 + }, + { + "epoch": 21.027955271565496, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 26327 + }, + { + "epoch": 21.028753993610223, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 26328 + }, + { + "epoch": 21.02955271565495, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 26329 + }, + { + "epoch": 21.03035143769968, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 26330 + }, + { + "epoch": 21.03115015974441, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 26331 + }, + { + "epoch": 21.031948881789138, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 26332 + }, + { + "epoch": 21.032747603833865, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26333 + }, + { + "epoch": 21.033546325878593, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 26334 + }, + { + "epoch": 21.034345047923324, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 26335 + }, + { + "epoch": 21.03514376996805, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 26336 + }, + { + "epoch": 21.03594249201278, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 26337 + }, + { + "epoch": 21.036741214057507, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 26338 + }, + { + "epoch": 21.037539936102238, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 26339 + }, + { + "epoch": 21.038338658146966, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 26340 + }, + { + "epoch": 21.039137380191693, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 26341 + }, + { + "epoch": 21.03993610223642, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26342 + }, + { + "epoch": 21.04073482428115, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 26343 + }, + { + "epoch": 21.04153354632588, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 26344 + }, + { + "epoch": 21.042332268370608, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 26345 + }, + { + "epoch": 21.043130990415335, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 26346 + }, + { + "epoch": 21.043929712460063, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 26347 + }, + { + "epoch": 21.044728434504794, + "grad_norm": 0.359375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 26348 + }, + { + "epoch": 21.04552715654952, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 26349 + }, + { + "epoch": 21.04632587859425, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 26350 + }, + { + "epoch": 21.047124600638977, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 26351 + }, + { + "epoch": 21.047923322683705, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 26352 + }, + { + "epoch": 21.048722044728436, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 26353 + }, + { + "epoch": 21.049520766773163, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 26354 + }, + { + "epoch": 21.05031948881789, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 26355 + }, + { + "epoch": 21.05111821086262, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 26356 + }, + { + "epoch": 21.05191693290735, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 26357 + }, + { + "epoch": 21.052715654952078, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 26358 + }, + { + "epoch": 21.053514376996805, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26359 + }, + { + "epoch": 21.054313099041533, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 26360 + }, + { + "epoch": 21.05511182108626, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 26361 + }, + { + "epoch": 21.05591054313099, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 26362 + }, + { + "epoch": 21.05670926517572, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 26363 + }, + { + "epoch": 21.057507987220447, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 26364 + }, + { + "epoch": 21.058306709265175, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26365 + }, + { + "epoch": 21.059105431309906, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26366 + }, + { + "epoch": 21.059904153354633, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 26367 + }, + { + "epoch": 21.06070287539936, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 26368 + }, + { + "epoch": 21.06150159744409, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 26369 + }, + { + "epoch": 21.062300319488816, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 26370 + }, + { + "epoch": 21.063099041533548, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 26371 + }, + { + "epoch": 21.063897763578275, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 26372 + }, + { + "epoch": 21.064696485623003, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 26373 + }, + { + "epoch": 21.06549520766773, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 26374 + }, + { + "epoch": 21.06629392971246, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26375 + }, + { + "epoch": 21.06709265175719, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 26376 + }, + { + "epoch": 21.067891373801917, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 26377 + }, + { + "epoch": 21.068690095846645, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 26378 + }, + { + "epoch": 21.069488817891372, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 26379 + }, + { + "epoch": 21.070287539936103, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 26380 + }, + { + "epoch": 21.07108626198083, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 26381 + }, + { + "epoch": 21.07188498402556, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 26382 + }, + { + "epoch": 21.072683706070286, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 26383 + }, + { + "epoch": 21.073482428115017, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 26384 + }, + { + "epoch": 21.074281150159745, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 26385 + }, + { + "epoch": 21.075079872204473, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26386 + }, + { + "epoch": 21.0758785942492, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 26387 + }, + { + "epoch": 21.076677316293928, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 26388 + }, + { + "epoch": 21.07747603833866, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 26389 + }, + { + "epoch": 21.078274760383387, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 26390 + }, + { + "epoch": 21.079073482428115, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 26391 + }, + { + "epoch": 21.079872204472842, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 26392 + }, + { + "epoch": 21.080670926517573, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 26393 + }, + { + "epoch": 21.0814696485623, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 26394 + }, + { + "epoch": 21.08226837060703, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 26395 + }, + { + "epoch": 21.083067092651756, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26396 + }, + { + "epoch": 21.083865814696484, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 26397 + }, + { + "epoch": 21.084664536741215, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26398 + }, + { + "epoch": 21.085463258785943, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 26399 + }, + { + "epoch": 21.08626198083067, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 26400 + }, + { + "epoch": 21.087060702875398, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 26401 + }, + { + "epoch": 21.08785942492013, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 26402 + }, + { + "epoch": 21.088658146964857, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.1047, + "step": 26403 + }, + { + "epoch": 21.089456869009584, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 26404 + }, + { + "epoch": 21.090255591054312, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26405 + }, + { + "epoch": 21.09105431309904, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 26406 + }, + { + "epoch": 21.09185303514377, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 26407 + }, + { + "epoch": 21.0926517571885, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 26408 + }, + { + "epoch": 21.093450479233226, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 26409 + }, + { + "epoch": 21.094249201277954, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 26410 + }, + { + "epoch": 21.095047923322685, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 26411 + }, + { + "epoch": 21.095846645367413, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 26412 + }, + { + "epoch": 21.09664536741214, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 26413 + }, + { + "epoch": 21.097444089456868, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 26414 + }, + { + "epoch": 21.0982428115016, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 26415 + }, + { + "epoch": 21.099041533546327, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 26416 + }, + { + "epoch": 21.099840255591054, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26417 + }, + { + "epoch": 21.100638977635782, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0752, + "step": 26418 + }, + { + "epoch": 21.10143769968051, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 26419 + }, + { + "epoch": 21.10223642172524, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 26420 + }, + { + "epoch": 21.10303514376997, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 26421 + }, + { + "epoch": 21.103833865814696, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26422 + }, + { + "epoch": 21.104632587859424, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26423 + }, + { + "epoch": 21.105431309904155, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 26424 + }, + { + "epoch": 21.106230031948883, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 26425 + }, + { + "epoch": 21.10702875399361, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 26426 + }, + { + "epoch": 21.107827476038338, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 26427 + }, + { + "epoch": 21.108626198083066, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 26428 + }, + { + "epoch": 21.109424920127797, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 26429 + }, + { + "epoch": 21.110223642172524, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26430 + }, + { + "epoch": 21.111022364217252, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 26431 + }, + { + "epoch": 21.11182108626198, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 26432 + }, + { + "epoch": 21.11261980830671, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 26433 + }, + { + "epoch": 21.11341853035144, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26434 + }, + { + "epoch": 21.114217252396166, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 26435 + }, + { + "epoch": 21.115015974440894, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26436 + }, + { + "epoch": 21.11581469648562, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 26437 + }, + { + "epoch": 21.116613418530353, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 26438 + }, + { + "epoch": 21.11741214057508, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 26439 + }, + { + "epoch": 21.118210862619808, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 26440 + }, + { + "epoch": 21.119009584664536, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 26441 + }, + { + "epoch": 21.119808306709267, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 26442 + }, + { + "epoch": 21.120607028753994, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 26443 + }, + { + "epoch": 21.121405750798722, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 26444 + }, + { + "epoch": 21.12220447284345, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 26445 + }, + { + "epoch": 21.123003194888177, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 26446 + }, + { + "epoch": 21.12380191693291, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 26447 + }, + { + "epoch": 21.124600638977636, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 26448 + }, + { + "epoch": 21.125399361022364, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 26449 + }, + { + "epoch": 21.12619808306709, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 26450 + }, + { + "epoch": 21.126996805111823, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 26451 + }, + { + "epoch": 21.12779552715655, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 26452 + }, + { + "epoch": 21.128594249201278, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26453 + }, + { + "epoch": 21.129392971246006, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 26454 + }, + { + "epoch": 21.130191693290733, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 26455 + }, + { + "epoch": 21.130990415335464, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 26456 + }, + { + "epoch": 21.131789137380192, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 26457 + }, + { + "epoch": 21.13258785942492, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 26458 + }, + { + "epoch": 21.133386581469647, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 26459 + }, + { + "epoch": 21.13418530351438, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26460 + }, + { + "epoch": 21.134984025559106, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 26461 + }, + { + "epoch": 21.135782747603834, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26462 + }, + { + "epoch": 21.13658146964856, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26463 + }, + { + "epoch": 21.13738019169329, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26464 + }, + { + "epoch": 21.13817891373802, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 26465 + }, + { + "epoch": 21.138977635782748, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 26466 + }, + { + "epoch": 21.139776357827476, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 26467 + }, + { + "epoch": 21.140575079872203, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 26468 + }, + { + "epoch": 21.141373801916934, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 26469 + }, + { + "epoch": 21.142172523961662, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 26470 + }, + { + "epoch": 21.14297124600639, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 26471 + }, + { + "epoch": 21.143769968051117, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 26472 + }, + { + "epoch": 21.144568690095845, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 26473 + }, + { + "epoch": 21.145367412140576, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 26474 + }, + { + "epoch": 21.146166134185304, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 26475 + }, + { + "epoch": 21.14696485623003, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 26476 + }, + { + "epoch": 21.14776357827476, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26477 + }, + { + "epoch": 21.14856230031949, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 26478 + }, + { + "epoch": 21.149361022364218, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0773, + "step": 26479 + }, + { + "epoch": 21.150159744408946, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 26480 + }, + { + "epoch": 21.150958466453673, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 26481 + }, + { + "epoch": 21.1517571884984, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 26482 + }, + { + "epoch": 21.152555910543132, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 26483 + }, + { + "epoch": 21.15335463258786, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 26484 + }, + { + "epoch": 21.154153354632587, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26485 + }, + { + "epoch": 21.154952076677315, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 26486 + }, + { + "epoch": 21.155750798722046, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 26487 + }, + { + "epoch": 21.156549520766774, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 26488 + }, + { + "epoch": 21.1573482428115, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 26489 + }, + { + "epoch": 21.15814696485623, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 26490 + }, + { + "epoch": 21.15894568690096, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 26491 + }, + { + "epoch": 21.159744408945688, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 26492 + }, + { + "epoch": 21.160543130990416, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 26493 + }, + { + "epoch": 21.161341853035143, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 26494 + }, + { + "epoch": 21.16214057507987, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 26495 + }, + { + "epoch": 21.162939297124602, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 26496 + }, + { + "epoch": 21.16373801916933, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 26497 + }, + { + "epoch": 21.164536741214057, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 26498 + }, + { + "epoch": 21.165335463258785, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 26499 + }, + { + "epoch": 21.166134185303516, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 26500 + }, + { + "epoch": 21.166932907348244, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 26501 + }, + { + "epoch": 21.16773162939297, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26502 + }, + { + "epoch": 21.1685303514377, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 26503 + }, + { + "epoch": 21.169329073482427, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26504 + }, + { + "epoch": 21.170127795527158, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 26505 + }, + { + "epoch": 21.170926517571885, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 26506 + }, + { + "epoch": 21.171725239616613, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 26507 + }, + { + "epoch": 21.17252396166134, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 26508 + }, + { + "epoch": 21.173322683706072, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 26509 + }, + { + "epoch": 21.1741214057508, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 26510 + }, + { + "epoch": 21.174920127795527, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 26511 + }, + { + "epoch": 21.175718849840255, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 26512 + }, + { + "epoch": 21.176517571884983, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 26513 + }, + { + "epoch": 21.177316293929714, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 26514 + }, + { + "epoch": 21.17811501597444, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 26515 + }, + { + "epoch": 21.17891373801917, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 26516 + }, + { + "epoch": 21.179712460063897, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 26517 + }, + { + "epoch": 21.180511182108628, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 26518 + }, + { + "epoch": 21.181309904153355, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 26519 + }, + { + "epoch": 21.182108626198083, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26520 + }, + { + "epoch": 21.18290734824281, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26521 + }, + { + "epoch": 21.18370607028754, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 26522 + }, + { + "epoch": 21.18450479233227, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 26523 + }, + { + "epoch": 21.185303514376997, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 26524 + }, + { + "epoch": 21.186102236421725, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 26525 + }, + { + "epoch": 21.186900958466452, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 26526 + }, + { + "epoch": 21.187699680511184, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 26527 + }, + { + "epoch": 21.18849840255591, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 26528 + }, + { + "epoch": 21.18929712460064, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 26529 + }, + { + "epoch": 21.190095846645367, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 26530 + }, + { + "epoch": 21.190894568690094, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 26531 + }, + { + "epoch": 21.191693290734825, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 26532 + }, + { + "epoch": 21.192492012779553, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 26533 + }, + { + "epoch": 21.19329073482428, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 26534 + }, + { + "epoch": 21.19408945686901, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 26535 + }, + { + "epoch": 21.19488817891374, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 26536 + }, + { + "epoch": 21.195686900958467, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 26537 + }, + { + "epoch": 21.196485623003195, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 26538 + }, + { + "epoch": 21.197284345047922, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 26539 + }, + { + "epoch": 21.19808306709265, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 26540 + }, + { + "epoch": 21.19888178913738, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 26541 + }, + { + "epoch": 21.19968051118211, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26542 + }, + { + "epoch": 21.200479233226837, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 26543 + }, + { + "epoch": 21.201277955271564, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 26544 + }, + { + "epoch": 21.202076677316295, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 26545 + }, + { + "epoch": 21.202875399361023, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 26546 + }, + { + "epoch": 21.20367412140575, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 26547 + }, + { + "epoch": 21.20447284345048, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 26548 + }, + { + "epoch": 21.205271565495206, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 26549 + }, + { + "epoch": 21.206070287539937, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26550 + }, + { + "epoch": 21.206869009584665, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 26551 + }, + { + "epoch": 21.207667731629392, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 26552 + }, + { + "epoch": 21.20846645367412, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 26553 + }, + { + "epoch": 21.20926517571885, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 26554 + }, + { + "epoch": 21.21006389776358, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 26555 + }, + { + "epoch": 21.210862619808307, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 26556 + }, + { + "epoch": 21.211661341853034, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 26557 + }, + { + "epoch": 21.212460063897762, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 26558 + }, + { + "epoch": 21.213258785942493, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 26559 + }, + { + "epoch": 21.21405750798722, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26560 + }, + { + "epoch": 21.21485623003195, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 26561 + }, + { + "epoch": 21.215654952076676, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 26562 + }, + { + "epoch": 21.216453674121407, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 26563 + }, + { + "epoch": 21.217252396166135, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 26564 + }, + { + "epoch": 21.218051118210862, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 26565 + }, + { + "epoch": 21.21884984025559, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 26566 + }, + { + "epoch": 21.21964856230032, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 26567 + }, + { + "epoch": 21.22044728434505, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26568 + }, + { + "epoch": 21.221246006389777, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 26569 + }, + { + "epoch": 21.222044728434504, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 26570 + }, + { + "epoch": 21.222843450479232, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 26571 + }, + { + "epoch": 21.223642172523963, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 26572 + }, + { + "epoch": 21.22444089456869, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 26573 + }, + { + "epoch": 21.22523961661342, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 26574 + }, + { + "epoch": 21.226038338658146, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 26575 + }, + { + "epoch": 21.226837060702877, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 26576 + }, + { + "epoch": 21.227635782747605, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 26577 + }, + { + "epoch": 21.228434504792332, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26578 + }, + { + "epoch": 21.22923322683706, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 26579 + }, + { + "epoch": 21.230031948881788, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 26580 + }, + { + "epoch": 21.23083067092652, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 26581 + }, + { + "epoch": 21.231629392971247, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26582 + }, + { + "epoch": 21.232428115015974, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 26583 + }, + { + "epoch": 21.233226837060702, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 26584 + }, + { + "epoch": 21.234025559105433, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 26585 + }, + { + "epoch": 21.23482428115016, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 26586 + }, + { + "epoch": 21.23562300319489, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26587 + }, + { + "epoch": 21.236421725239616, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 26588 + }, + { + "epoch": 21.237220447284344, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 26589 + }, + { + "epoch": 21.238019169329075, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 26590 + }, + { + "epoch": 21.238817891373802, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 26591 + }, + { + "epoch": 21.23961661341853, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26592 + }, + { + "epoch": 21.240415335463258, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 26593 + }, + { + "epoch": 21.24121405750799, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 26594 + }, + { + "epoch": 21.242012779552716, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26595 + }, + { + "epoch": 21.242811501597444, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 26596 + }, + { + "epoch": 21.24361022364217, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 26597 + }, + { + "epoch": 21.2444089456869, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 26598 + }, + { + "epoch": 21.24520766773163, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 26599 + }, + { + "epoch": 21.24600638977636, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 26600 + }, + { + "epoch": 21.246805111821086, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 26601 + }, + { + "epoch": 21.247603833865814, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 26602 + }, + { + "epoch": 21.248402555910545, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 26603 + }, + { + "epoch": 21.249201277955272, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26604 + }, + { + "epoch": 21.25, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 26605 + }, + { + "epoch": 21.250798722044728, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 26606 + }, + { + "epoch": 21.251597444089455, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 26607 + }, + { + "epoch": 21.252396166134186, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 26608 + }, + { + "epoch": 21.253194888178914, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 26609 + }, + { + "epoch": 21.25399361022364, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 26610 + }, + { + "epoch": 21.25479233226837, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 26611 + }, + { + "epoch": 21.2555910543131, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 26612 + }, + { + "epoch": 21.25638977635783, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 26613 + }, + { + "epoch": 21.257188498402556, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 26614 + }, + { + "epoch": 21.257987220447284, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 26615 + }, + { + "epoch": 21.25878594249201, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 26616 + }, + { + "epoch": 21.259584664536742, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 26617 + }, + { + "epoch": 21.26038338658147, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 26618 + }, + { + "epoch": 21.261182108626198, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 26619 + }, + { + "epoch": 21.261980830670925, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 26620 + }, + { + "epoch": 21.262779552715656, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 26621 + }, + { + "epoch": 21.263578274760384, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 26622 + }, + { + "epoch": 21.26437699680511, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 26623 + }, + { + "epoch": 21.26517571884984, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 26624 + }, + { + "epoch": 21.265974440894567, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 26625 + }, + { + "epoch": 21.266773162939298, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 26626 + }, + { + "epoch": 21.267571884984026, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 26627 + }, + { + "epoch": 21.268370607028753, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 26628 + }, + { + "epoch": 21.26916932907348, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 26629 + }, + { + "epoch": 21.269968051118212, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26630 + }, + { + "epoch": 21.27076677316294, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 26631 + }, + { + "epoch": 21.271565495207668, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 26632 + }, + { + "epoch": 21.272364217252395, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 26633 + }, + { + "epoch": 21.273162939297123, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 26634 + }, + { + "epoch": 21.273961661341854, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 26635 + }, + { + "epoch": 21.27476038338658, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 26636 + }, + { + "epoch": 21.27555910543131, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 26637 + }, + { + "epoch": 21.276357827476037, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 26638 + }, + { + "epoch": 21.277156549520768, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 26639 + }, + { + "epoch": 21.277955271565496, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 26640 + }, + { + "epoch": 21.278753993610223, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 26641 + }, + { + "epoch": 21.27955271565495, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26642 + }, + { + "epoch": 21.28035143769968, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 26643 + }, + { + "epoch": 21.28115015974441, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 26644 + }, + { + "epoch": 21.281948881789138, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 26645 + }, + { + "epoch": 21.282747603833865, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 26646 + }, + { + "epoch": 21.283546325878593, + "grad_norm": 0.09619140625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 26647 + }, + { + "epoch": 21.284345047923324, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 26648 + }, + { + "epoch": 21.28514376996805, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 26649 + }, + { + "epoch": 21.28594249201278, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 26650 + }, + { + "epoch": 21.286741214057507, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 26651 + }, + { + "epoch": 21.287539936102238, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 26652 + }, + { + "epoch": 21.288338658146966, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 26653 + }, + { + "epoch": 21.289137380191693, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 26654 + }, + { + "epoch": 21.28993610223642, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 26655 + }, + { + "epoch": 21.29073482428115, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 26656 + }, + { + "epoch": 21.29153354632588, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 26657 + }, + { + "epoch": 21.292332268370608, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 26658 + }, + { + "epoch": 21.293130990415335, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 26659 + }, + { + "epoch": 21.293929712460063, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 26660 + }, + { + "epoch": 21.294728434504794, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 26661 + }, + { + "epoch": 21.29552715654952, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 26662 + }, + { + "epoch": 21.29632587859425, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 26663 + }, + { + "epoch": 21.297124600638977, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 26664 + }, + { + "epoch": 21.297923322683705, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26665 + }, + { + "epoch": 21.298722044728436, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 26666 + }, + { + "epoch": 21.299520766773163, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26667 + }, + { + "epoch": 21.30031948881789, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 26668 + }, + { + "epoch": 21.30111821086262, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 26669 + }, + { + "epoch": 21.30191693290735, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 26670 + }, + { + "epoch": 21.302715654952078, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 26671 + }, + { + "epoch": 21.303514376996805, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 26672 + }, + { + "epoch": 21.304313099041533, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 26673 + }, + { + "epoch": 21.30511182108626, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26674 + }, + { + "epoch": 21.30591054313099, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 26675 + }, + { + "epoch": 21.30670926517572, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 26676 + }, + { + "epoch": 21.307507987220447, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26677 + }, + { + "epoch": 21.308306709265175, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 26678 + }, + { + "epoch": 21.309105431309906, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 26679 + }, + { + "epoch": 21.309904153354633, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 26680 + }, + { + "epoch": 21.31070287539936, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 26681 + }, + { + "epoch": 21.31150159744409, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 26682 + }, + { + "epoch": 21.312300319488816, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 26683 + }, + { + "epoch": 21.313099041533548, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 26684 + }, + { + "epoch": 21.313897763578275, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 26685 + }, + { + "epoch": 21.314696485623003, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 26686 + }, + { + "epoch": 21.31549520766773, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 26687 + }, + { + "epoch": 21.31629392971246, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 26688 + }, + { + "epoch": 21.31709265175719, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 26689 + }, + { + "epoch": 21.317891373801917, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 26690 + }, + { + "epoch": 21.318690095846645, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 26691 + }, + { + "epoch": 21.319488817891372, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 26692 + }, + { + "epoch": 21.320287539936103, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 26693 + }, + { + "epoch": 21.32108626198083, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 26694 + }, + { + "epoch": 21.32188498402556, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 26695 + }, + { + "epoch": 21.322683706070286, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 26696 + }, + { + "epoch": 21.323482428115017, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 26697 + }, + { + "epoch": 21.324281150159745, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 26698 + }, + { + "epoch": 21.325079872204473, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 26699 + }, + { + "epoch": 21.3258785942492, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 26700 + }, + { + "epoch": 21.326677316293928, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 26701 + }, + { + "epoch": 21.32747603833866, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 26702 + }, + { + "epoch": 21.328274760383387, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 26703 + }, + { + "epoch": 21.329073482428115, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 26704 + }, + { + "epoch": 21.329872204472842, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 26705 + }, + { + "epoch": 21.330670926517573, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 26706 + }, + { + "epoch": 21.3314696485623, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 26707 + }, + { + "epoch": 21.33226837060703, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26708 + }, + { + "epoch": 21.333067092651756, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26709 + }, + { + "epoch": 21.333865814696484, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26710 + }, + { + "epoch": 21.334664536741215, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 26711 + }, + { + "epoch": 21.335463258785943, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 26712 + }, + { + "epoch": 21.33626198083067, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26713 + }, + { + "epoch": 21.337060702875398, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26714 + }, + { + "epoch": 21.33785942492013, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 26715 + }, + { + "epoch": 21.338658146964857, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 26716 + }, + { + "epoch": 21.339456869009584, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 26717 + }, + { + "epoch": 21.340255591054312, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 26718 + }, + { + "epoch": 21.341054313099043, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26719 + }, + { + "epoch": 21.34185303514377, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 26720 + }, + { + "epoch": 21.3426517571885, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 26721 + }, + { + "epoch": 21.343450479233226, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 26722 + }, + { + "epoch": 21.344249201277954, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26723 + }, + { + "epoch": 21.345047923322685, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 26724 + }, + { + "epoch": 21.345846645367413, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 26725 + }, + { + "epoch": 21.34664536741214, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 26726 + }, + { + "epoch": 21.347444089456868, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 26727 + }, + { + "epoch": 21.3482428115016, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 26728 + }, + { + "epoch": 21.349041533546327, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 26729 + }, + { + "epoch": 21.349840255591054, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 26730 + }, + { + "epoch": 21.350638977635782, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 26731 + }, + { + "epoch": 21.35143769968051, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 26732 + }, + { + "epoch": 21.35223642172524, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 26733 + }, + { + "epoch": 21.35303514376997, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 26734 + }, + { + "epoch": 21.353833865814696, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 26735 + }, + { + "epoch": 21.354632587859424, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 26736 + }, + { + "epoch": 21.355431309904155, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 26737 + }, + { + "epoch": 21.356230031948883, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 26738 + }, + { + "epoch": 21.35702875399361, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 26739 + }, + { + "epoch": 21.357827476038338, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 26740 + }, + { + "epoch": 21.358626198083066, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 26741 + }, + { + "epoch": 21.359424920127797, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 26742 + }, + { + "epoch": 21.360223642172524, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 26743 + }, + { + "epoch": 21.361022364217252, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 26744 + }, + { + "epoch": 21.36182108626198, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 26745 + }, + { + "epoch": 21.36261980830671, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 26746 + }, + { + "epoch": 21.36341853035144, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 26747 + }, + { + "epoch": 21.364217252396166, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 26748 + }, + { + "epoch": 21.365015974440894, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 26749 + }, + { + "epoch": 21.36581469648562, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 26750 + }, + { + "epoch": 21.366613418530353, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 26751 + }, + { + "epoch": 21.36741214057508, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 26752 + }, + { + "epoch": 21.368210862619808, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 26753 + }, + { + "epoch": 21.369009584664536, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 26754 + }, + { + "epoch": 21.369808306709267, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 26755 + }, + { + "epoch": 21.370607028753994, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26756 + }, + { + "epoch": 21.371405750798722, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 26757 + }, + { + "epoch": 21.37220447284345, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 26758 + }, + { + "epoch": 21.373003194888177, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 26759 + }, + { + "epoch": 21.37380191693291, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 26760 + }, + { + "epoch": 21.374600638977636, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 26761 + }, + { + "epoch": 21.375399361022364, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 26762 + }, + { + "epoch": 21.37619808306709, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 26763 + }, + { + "epoch": 21.376996805111823, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26764 + }, + { + "epoch": 21.37779552715655, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 26765 + }, + { + "epoch": 21.378594249201278, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 26766 + }, + { + "epoch": 21.379392971246006, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 26767 + }, + { + "epoch": 21.380191693290733, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 26768 + }, + { + "epoch": 21.380990415335464, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 26769 + }, + { + "epoch": 21.381789137380192, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 26770 + }, + { + "epoch": 21.38258785942492, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 26771 + }, + { + "epoch": 21.383386581469647, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 26772 + }, + { + "epoch": 21.38418530351438, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 26773 + }, + { + "epoch": 21.384984025559106, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 26774 + }, + { + "epoch": 21.385782747603834, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 26775 + }, + { + "epoch": 21.38658146964856, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 26776 + }, + { + "epoch": 21.38738019169329, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 26777 + }, + { + "epoch": 21.38817891373802, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 26778 + }, + { + "epoch": 21.388977635782748, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 26779 + }, + { + "epoch": 21.389776357827476, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 26780 + }, + { + "epoch": 21.390575079872203, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 26781 + }, + { + "epoch": 21.391373801916934, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 26782 + }, + { + "epoch": 21.392172523961662, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 26783 + }, + { + "epoch": 21.39297124600639, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 26784 + }, + { + "epoch": 21.393769968051117, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 26785 + }, + { + "epoch": 21.394568690095845, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 26786 + }, + { + "epoch": 21.395367412140576, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 26787 + }, + { + "epoch": 21.396166134185304, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 26788 + }, + { + "epoch": 21.39696485623003, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 26789 + }, + { + "epoch": 21.39776357827476, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 26790 + }, + { + "epoch": 21.39856230031949, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 26791 + }, + { + "epoch": 21.399361022364218, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 26792 + }, + { + "epoch": 21.400159744408946, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26793 + }, + { + "epoch": 21.400958466453673, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 26794 + }, + { + "epoch": 21.4017571884984, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 26795 + }, + { + "epoch": 21.402555910543132, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 26796 + }, + { + "epoch": 21.40335463258786, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 26797 + }, + { + "epoch": 21.404153354632587, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 26798 + }, + { + "epoch": 21.404952076677315, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 26799 + }, + { + "epoch": 21.405750798722046, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 26800 + }, + { + "epoch": 21.406549520766774, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26801 + }, + { + "epoch": 21.4073482428115, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 26802 + }, + { + "epoch": 21.40814696485623, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 26803 + }, + { + "epoch": 21.408945686900957, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 26804 + }, + { + "epoch": 21.409744408945688, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 26805 + }, + { + "epoch": 21.410543130990416, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 26806 + }, + { + "epoch": 21.411341853035143, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 26807 + }, + { + "epoch": 21.41214057507987, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 26808 + }, + { + "epoch": 21.412939297124602, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 26809 + }, + { + "epoch": 21.41373801916933, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 26810 + }, + { + "epoch": 21.414536741214057, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 26811 + }, + { + "epoch": 21.415335463258785, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 26812 + }, + { + "epoch": 21.416134185303516, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 26813 + }, + { + "epoch": 21.416932907348244, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 26814 + }, + { + "epoch": 21.41773162939297, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 26815 + }, + { + "epoch": 21.4185303514377, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 26816 + }, + { + "epoch": 21.419329073482427, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 26817 + }, + { + "epoch": 21.420127795527158, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 26818 + }, + { + "epoch": 21.420926517571885, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 26819 + }, + { + "epoch": 21.421725239616613, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26820 + }, + { + "epoch": 21.42252396166134, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 26821 + }, + { + "epoch": 21.423322683706072, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 26822 + }, + { + "epoch": 21.4241214057508, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 26823 + }, + { + "epoch": 21.424920127795527, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 26824 + }, + { + "epoch": 21.425718849840255, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 26825 + }, + { + "epoch": 21.426517571884983, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 26826 + }, + { + "epoch": 21.427316293929714, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 26827 + }, + { + "epoch": 21.42811501597444, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 26828 + }, + { + "epoch": 21.42891373801917, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 26829 + }, + { + "epoch": 21.429712460063897, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 26830 + }, + { + "epoch": 21.430511182108628, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 26831 + }, + { + "epoch": 21.431309904153355, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 26832 + }, + { + "epoch": 21.432108626198083, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 26833 + }, + { + "epoch": 21.43290734824281, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 26834 + }, + { + "epoch": 21.43370607028754, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 26835 + }, + { + "epoch": 21.43450479233227, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 26836 + }, + { + "epoch": 21.435303514376997, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 26837 + }, + { + "epoch": 21.436102236421725, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 26838 + }, + { + "epoch": 21.436900958466452, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 26839 + }, + { + "epoch": 21.437699680511184, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 26840 + }, + { + "epoch": 21.43849840255591, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 26841 + }, + { + "epoch": 21.43929712460064, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 26842 + }, + { + "epoch": 21.440095846645367, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 26843 + }, + { + "epoch": 21.440894568690094, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 26844 + }, + { + "epoch": 21.441693290734825, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26845 + }, + { + "epoch": 21.442492012779553, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 26846 + }, + { + "epoch": 21.44329073482428, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 26847 + }, + { + "epoch": 21.44408945686901, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 26848 + }, + { + "epoch": 21.44488817891374, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 26849 + }, + { + "epoch": 21.445686900958467, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 26850 + }, + { + "epoch": 21.446485623003195, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 26851 + }, + { + "epoch": 21.447284345047922, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 26852 + }, + { + "epoch": 21.44808306709265, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 26853 + }, + { + "epoch": 21.44888178913738, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 26854 + }, + { + "epoch": 21.44968051118211, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 26855 + }, + { + "epoch": 21.450479233226837, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 26856 + }, + { + "epoch": 21.451277955271564, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 26857 + }, + { + "epoch": 21.452076677316295, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 26858 + }, + { + "epoch": 21.452875399361023, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 26859 + }, + { + "epoch": 21.45367412140575, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 26860 + }, + { + "epoch": 21.45447284345048, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 26861 + }, + { + "epoch": 21.455271565495206, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 26862 + }, + { + "epoch": 21.456070287539937, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0761, + "step": 26863 + }, + { + "epoch": 21.456869009584665, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 26864 + }, + { + "epoch": 21.457667731629392, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 26865 + }, + { + "epoch": 21.45846645367412, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 26866 + }, + { + "epoch": 21.45926517571885, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1064, + "step": 26867 + }, + { + "epoch": 21.46006389776358, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 26868 + }, + { + "epoch": 21.460862619808307, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26869 + }, + { + "epoch": 21.461661341853034, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 26870 + }, + { + "epoch": 21.462460063897762, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 26871 + }, + { + "epoch": 21.463258785942493, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26872 + }, + { + "epoch": 21.46405750798722, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 26873 + }, + { + "epoch": 21.46485623003195, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 26874 + }, + { + "epoch": 21.465654952076676, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 26875 + }, + { + "epoch": 21.466453674121407, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 26876 + }, + { + "epoch": 21.467252396166135, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 26877 + }, + { + "epoch": 21.468051118210862, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 26878 + }, + { + "epoch": 21.46884984025559, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 26879 + }, + { + "epoch": 21.46964856230032, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 26880 + }, + { + "epoch": 21.47044728434505, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 26881 + }, + { + "epoch": 21.471246006389777, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26882 + }, + { + "epoch": 21.472044728434504, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 26883 + }, + { + "epoch": 21.472843450479232, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 26884 + }, + { + "epoch": 21.473642172523963, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 26885 + }, + { + "epoch": 21.47444089456869, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 26886 + }, + { + "epoch": 21.47523961661342, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 26887 + }, + { + "epoch": 21.476038338658146, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 26888 + }, + { + "epoch": 21.476837060702877, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 26889 + }, + { + "epoch": 21.477635782747605, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 26890 + }, + { + "epoch": 21.478434504792332, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26891 + }, + { + "epoch": 21.47923322683706, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26892 + }, + { + "epoch": 21.480031948881788, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 26893 + }, + { + "epoch": 21.48083067092652, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 26894 + }, + { + "epoch": 21.481629392971247, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 26895 + }, + { + "epoch": 21.482428115015974, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 26896 + }, + { + "epoch": 21.483226837060702, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 26897 + }, + { + "epoch": 21.484025559105433, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 26898 + }, + { + "epoch": 21.48482428115016, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 26899 + }, + { + "epoch": 21.48562300319489, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 26900 + }, + { + "epoch": 21.486421725239616, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 26901 + }, + { + "epoch": 21.487220447284344, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 26902 + }, + { + "epoch": 21.488019169329075, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 26903 + }, + { + "epoch": 21.488817891373802, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 26904 + }, + { + "epoch": 21.48961661341853, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 26905 + }, + { + "epoch": 21.490415335463258, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 26906 + }, + { + "epoch": 21.49121405750799, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26907 + }, + { + "epoch": 21.492012779552716, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 26908 + }, + { + "epoch": 21.492811501597444, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 26909 + }, + { + "epoch": 21.49361022364217, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 26910 + }, + { + "epoch": 21.4944089456869, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 26911 + }, + { + "epoch": 21.49520766773163, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 26912 + }, + { + "epoch": 21.49600638977636, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 26913 + }, + { + "epoch": 21.496805111821086, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 26914 + }, + { + "epoch": 21.497603833865814, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 26915 + }, + { + "epoch": 21.498402555910545, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 26916 + }, + { + "epoch": 21.499201277955272, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 26917 + }, + { + "epoch": 21.5, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 26918 + }, + { + "epoch": 21.500798722044728, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 26919 + }, + { + "epoch": 21.501597444089455, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 26920 + }, + { + "epoch": 21.502396166134186, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 26921 + }, + { + "epoch": 21.503194888178914, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 26922 + }, + { + "epoch": 21.50399361022364, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 26923 + }, + { + "epoch": 21.50479233226837, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 26924 + }, + { + "epoch": 21.5055910543131, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 26925 + }, + { + "epoch": 21.50638977635783, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 26926 + }, + { + "epoch": 21.507188498402556, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 26927 + }, + { + "epoch": 21.507987220447284, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 26928 + }, + { + "epoch": 21.50878594249201, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 26929 + }, + { + "epoch": 21.509584664536742, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 26930 + }, + { + "epoch": 21.51038338658147, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 26931 + }, + { + "epoch": 21.511182108626198, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 26932 + }, + { + "epoch": 21.511980830670925, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 26933 + }, + { + "epoch": 21.512779552715656, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 26934 + }, + { + "epoch": 21.513578274760384, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 26935 + }, + { + "epoch": 21.51437699680511, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 26936 + }, + { + "epoch": 21.51517571884984, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 26937 + }, + { + "epoch": 21.515974440894567, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 26938 + }, + { + "epoch": 21.516773162939298, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 26939 + }, + { + "epoch": 21.517571884984026, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 26940 + }, + { + "epoch": 21.518370607028753, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 26941 + }, + { + "epoch": 21.51916932907348, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 26942 + }, + { + "epoch": 21.519968051118212, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 26943 + }, + { + "epoch": 21.52076677316294, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 26944 + }, + { + "epoch": 21.521565495207668, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 26945 + }, + { + "epoch": 21.522364217252395, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 26946 + }, + { + "epoch": 21.523162939297123, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 26947 + }, + { + "epoch": 21.523961661341854, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 26948 + }, + { + "epoch": 21.52476038338658, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 26949 + }, + { + "epoch": 21.52555910543131, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 26950 + }, + { + "epoch": 21.526357827476037, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 26951 + }, + { + "epoch": 21.527156549520768, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 26952 + }, + { + "epoch": 21.527955271565496, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 26953 + }, + { + "epoch": 21.528753993610223, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 26954 + }, + { + "epoch": 21.52955271565495, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 26955 + }, + { + "epoch": 21.53035143769968, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 26956 + }, + { + "epoch": 21.53115015974441, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 26957 + }, + { + "epoch": 21.531948881789138, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 26958 + }, + { + "epoch": 21.532747603833865, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 26959 + }, + { + "epoch": 21.533546325878593, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 26960 + }, + { + "epoch": 21.534345047923324, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 26961 + }, + { + "epoch": 21.53514376996805, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 26962 + }, + { + "epoch": 21.53594249201278, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 26963 + }, + { + "epoch": 21.536741214057507, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 26964 + }, + { + "epoch": 21.537539936102235, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 26965 + }, + { + "epoch": 21.538338658146966, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 26966 + }, + { + "epoch": 21.539137380191693, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 26967 + }, + { + "epoch": 21.53993610223642, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 26968 + }, + { + "epoch": 21.54073482428115, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 26969 + }, + { + "epoch": 21.54153354632588, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 26970 + }, + { + "epoch": 21.542332268370608, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1088, + "step": 26971 + }, + { + "epoch": 21.543130990415335, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 26972 + }, + { + "epoch": 21.543929712460063, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 26973 + }, + { + "epoch": 21.544728434504794, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 26974 + }, + { + "epoch": 21.54552715654952, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 26975 + }, + { + "epoch": 21.54632587859425, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 26976 + }, + { + "epoch": 21.547124600638977, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 26977 + }, + { + "epoch": 21.547923322683705, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 26978 + }, + { + "epoch": 21.548722044728436, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 26979 + }, + { + "epoch": 21.549520766773163, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 26980 + }, + { + "epoch": 21.55031948881789, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 26981 + }, + { + "epoch": 21.55111821086262, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 26982 + }, + { + "epoch": 21.55191693290735, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 26983 + }, + { + "epoch": 21.552715654952078, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 26984 + }, + { + "epoch": 21.553514376996805, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 26985 + }, + { + "epoch": 21.554313099041533, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 26986 + }, + { + "epoch": 21.55511182108626, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 26987 + }, + { + "epoch": 21.55591054313099, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 26988 + }, + { + "epoch": 21.55670926517572, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 26989 + }, + { + "epoch": 21.557507987220447, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 26990 + }, + { + "epoch": 21.558306709265175, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 26991 + }, + { + "epoch": 21.559105431309906, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 26992 + }, + { + "epoch": 21.559904153354633, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 26993 + }, + { + "epoch": 21.56070287539936, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 26994 + }, + { + "epoch": 21.56150159744409, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 26995 + }, + { + "epoch": 21.562300319488816, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 26996 + }, + { + "epoch": 21.563099041533548, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 26997 + }, + { + "epoch": 21.563897763578275, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 26998 + }, + { + "epoch": 21.564696485623003, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 26999 + }, + { + "epoch": 21.56549520766773, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 27000 + }, + { + "epoch": 21.56629392971246, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 27001 + }, + { + "epoch": 21.56709265175719, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 27002 + }, + { + "epoch": 21.567891373801917, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 27003 + }, + { + "epoch": 21.568690095846645, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 27004 + }, + { + "epoch": 21.569488817891372, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 27005 + }, + { + "epoch": 21.570287539936103, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 27006 + }, + { + "epoch": 21.57108626198083, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 27007 + }, + { + "epoch": 21.57188498402556, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 27008 + }, + { + "epoch": 21.572683706070286, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 27009 + }, + { + "epoch": 21.573482428115017, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 27010 + }, + { + "epoch": 21.574281150159745, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 27011 + }, + { + "epoch": 21.575079872204473, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 27012 + }, + { + "epoch": 21.5758785942492, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 27013 + }, + { + "epoch": 21.576677316293928, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27014 + }, + { + "epoch": 21.57747603833866, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 27015 + }, + { + "epoch": 21.578274760383387, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27016 + }, + { + "epoch": 21.579073482428115, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 27017 + }, + { + "epoch": 21.579872204472842, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 27018 + }, + { + "epoch": 21.580670926517573, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 27019 + }, + { + "epoch": 21.5814696485623, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 27020 + }, + { + "epoch": 21.58226837060703, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 27021 + }, + { + "epoch": 21.583067092651756, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 27022 + }, + { + "epoch": 21.583865814696484, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 27023 + }, + { + "epoch": 21.584664536741215, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 27024 + }, + { + "epoch": 21.585463258785943, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 27025 + }, + { + "epoch": 21.58626198083067, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 27026 + }, + { + "epoch": 21.587060702875398, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 27027 + }, + { + "epoch": 21.58785942492013, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 27028 + }, + { + "epoch": 21.588658146964857, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27029 + }, + { + "epoch": 21.589456869009584, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 27030 + }, + { + "epoch": 21.590255591054312, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 27031 + }, + { + "epoch": 21.591054313099043, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 27032 + }, + { + "epoch": 21.59185303514377, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 27033 + }, + { + "epoch": 21.5926517571885, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 27034 + }, + { + "epoch": 21.593450479233226, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 27035 + }, + { + "epoch": 21.594249201277954, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 27036 + }, + { + "epoch": 21.595047923322685, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 27037 + }, + { + "epoch": 21.595846645367413, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 27038 + }, + { + "epoch": 21.59664536741214, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 27039 + }, + { + "epoch": 21.597444089456868, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 27040 + }, + { + "epoch": 21.5982428115016, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 27041 + }, + { + "epoch": 21.599041533546327, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 27042 + }, + { + "epoch": 21.599840255591054, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 27043 + }, + { + "epoch": 21.600638977635782, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 27044 + }, + { + "epoch": 21.60143769968051, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 27045 + }, + { + "epoch": 21.60223642172524, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 27046 + }, + { + "epoch": 21.60303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 27047 + }, + { + "epoch": 21.603833865814696, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 27048 + }, + { + "epoch": 21.604632587859424, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27049 + }, + { + "epoch": 21.605431309904155, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 27050 + }, + { + "epoch": 21.606230031948883, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 27051 + }, + { + "epoch": 21.60702875399361, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 27052 + }, + { + "epoch": 21.607827476038338, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 27053 + }, + { + "epoch": 21.608626198083066, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 27054 + }, + { + "epoch": 21.609424920127797, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 27055 + }, + { + "epoch": 21.610223642172524, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 27056 + }, + { + "epoch": 21.611022364217252, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 27057 + }, + { + "epoch": 21.61182108626198, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 27058 + }, + { + "epoch": 21.61261980830671, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 27059 + }, + { + "epoch": 21.61341853035144, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27060 + }, + { + "epoch": 21.614217252396166, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 27061 + }, + { + "epoch": 21.615015974440894, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 27062 + }, + { + "epoch": 21.61581469648562, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 27063 + }, + { + "epoch": 21.616613418530353, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 27064 + }, + { + "epoch": 21.61741214057508, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 27065 + }, + { + "epoch": 21.618210862619808, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 27066 + }, + { + "epoch": 21.619009584664536, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 27067 + }, + { + "epoch": 21.619808306709267, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 27068 + }, + { + "epoch": 21.620607028753994, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 27069 + }, + { + "epoch": 21.621405750798722, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 27070 + }, + { + "epoch": 21.62220447284345, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 27071 + }, + { + "epoch": 21.623003194888177, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27072 + }, + { + "epoch": 21.62380191693291, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 27073 + }, + { + "epoch": 21.624600638977636, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 27074 + }, + { + "epoch": 21.625399361022364, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 27075 + }, + { + "epoch": 21.62619808306709, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 27076 + }, + { + "epoch": 21.626996805111823, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 27077 + }, + { + "epoch": 21.62779552715655, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 27078 + }, + { + "epoch": 21.628594249201278, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 27079 + }, + { + "epoch": 21.629392971246006, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 27080 + }, + { + "epoch": 21.630191693290733, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 27081 + }, + { + "epoch": 21.630990415335464, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 27082 + }, + { + "epoch": 21.631789137380192, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27083 + }, + { + "epoch": 21.63258785942492, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 27084 + }, + { + "epoch": 21.633386581469647, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 27085 + }, + { + "epoch": 21.63418530351438, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 27086 + }, + { + "epoch": 21.634984025559106, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 27087 + }, + { + "epoch": 21.635782747603834, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 27088 + }, + { + "epoch": 21.63658146964856, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 27089 + }, + { + "epoch": 21.63738019169329, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 27090 + }, + { + "epoch": 21.63817891373802, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 27091 + }, + { + "epoch": 21.638977635782748, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 27092 + }, + { + "epoch": 21.639776357827476, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 27093 + }, + { + "epoch": 21.640575079872203, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 27094 + }, + { + "epoch": 21.641373801916934, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 27095 + }, + { + "epoch": 21.642172523961662, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 27096 + }, + { + "epoch": 21.64297124600639, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 27097 + }, + { + "epoch": 21.643769968051117, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 27098 + }, + { + "epoch": 21.644568690095845, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 27099 + }, + { + "epoch": 21.645367412140576, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 27100 + }, + { + "epoch": 21.646166134185304, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 27101 + }, + { + "epoch": 21.64696485623003, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 27102 + }, + { + "epoch": 21.64776357827476, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 27103 + }, + { + "epoch": 21.64856230031949, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 27104 + }, + { + "epoch": 21.649361022364218, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 27105 + }, + { + "epoch": 21.650159744408946, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 27106 + }, + { + "epoch": 21.650958466453673, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27107 + }, + { + "epoch": 21.6517571884984, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 27108 + }, + { + "epoch": 21.652555910543132, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 27109 + }, + { + "epoch": 21.65335463258786, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 27110 + }, + { + "epoch": 21.654153354632587, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 27111 + }, + { + "epoch": 21.654952076677315, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27112 + }, + { + "epoch": 21.655750798722046, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 27113 + }, + { + "epoch": 21.656549520766774, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 27114 + }, + { + "epoch": 21.6573482428115, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 27115 + }, + { + "epoch": 21.65814696485623, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 27116 + }, + { + "epoch": 21.658945686900957, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 27117 + }, + { + "epoch": 21.659744408945688, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 27118 + }, + { + "epoch": 21.660543130990416, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 27119 + }, + { + "epoch": 21.661341853035143, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27120 + }, + { + "epoch": 21.66214057507987, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 27121 + }, + { + "epoch": 21.662939297124602, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 27122 + }, + { + "epoch": 21.66373801916933, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 27123 + }, + { + "epoch": 21.664536741214057, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 27124 + }, + { + "epoch": 21.665335463258785, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 27125 + }, + { + "epoch": 21.666134185303516, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 27126 + }, + { + "epoch": 21.666932907348244, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 27127 + }, + { + "epoch": 21.66773162939297, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 27128 + }, + { + "epoch": 21.6685303514377, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 27129 + }, + { + "epoch": 21.669329073482427, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 27130 + }, + { + "epoch": 21.670127795527158, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 27131 + }, + { + "epoch": 21.670926517571885, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 27132 + }, + { + "epoch": 21.671725239616613, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 27133 + }, + { + "epoch": 21.67252396166134, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 27134 + }, + { + "epoch": 21.673322683706072, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 27135 + }, + { + "epoch": 21.6741214057508, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 27136 + }, + { + "epoch": 21.674920127795527, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 27137 + }, + { + "epoch": 21.675718849840255, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 27138 + }, + { + "epoch": 21.676517571884983, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 27139 + }, + { + "epoch": 21.677316293929714, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 27140 + }, + { + "epoch": 21.67811501597444, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 27141 + }, + { + "epoch": 21.67891373801917, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 27142 + }, + { + "epoch": 21.679712460063897, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 27143 + }, + { + "epoch": 21.680511182108628, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27144 + }, + { + "epoch": 21.681309904153355, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 27145 + }, + { + "epoch": 21.682108626198083, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 27146 + }, + { + "epoch": 21.68290734824281, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 27147 + }, + { + "epoch": 21.68370607028754, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 27148 + }, + { + "epoch": 21.68450479233227, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27149 + }, + { + "epoch": 21.685303514376997, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 27150 + }, + { + "epoch": 21.686102236421725, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 27151 + }, + { + "epoch": 21.686900958466452, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 27152 + }, + { + "epoch": 21.687699680511184, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 27153 + }, + { + "epoch": 21.68849840255591, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 27154 + }, + { + "epoch": 21.68929712460064, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0763, + "step": 27155 + }, + { + "epoch": 21.690095846645367, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 27156 + }, + { + "epoch": 21.690894568690094, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 27157 + }, + { + "epoch": 21.691693290734825, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 27158 + }, + { + "epoch": 21.692492012779553, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 27159 + }, + { + "epoch": 21.69329073482428, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 27160 + }, + { + "epoch": 21.69408945686901, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 27161 + }, + { + "epoch": 21.69488817891374, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 27162 + }, + { + "epoch": 21.695686900958467, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 27163 + }, + { + "epoch": 21.696485623003195, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 27164 + }, + { + "epoch": 21.697284345047922, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 27165 + }, + { + "epoch": 21.69808306709265, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 27166 + }, + { + "epoch": 21.69888178913738, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 27167 + }, + { + "epoch": 21.69968051118211, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 27168 + }, + { + "epoch": 21.700479233226837, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 27169 + }, + { + "epoch": 21.701277955271564, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 27170 + }, + { + "epoch": 21.702076677316295, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 27171 + }, + { + "epoch": 21.702875399361023, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 27172 + }, + { + "epoch": 21.70367412140575, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 27173 + }, + { + "epoch": 21.70447284345048, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 27174 + }, + { + "epoch": 21.705271565495206, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27175 + }, + { + "epoch": 21.706070287539937, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 27176 + }, + { + "epoch": 21.706869009584665, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27177 + }, + { + "epoch": 21.707667731629392, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 27178 + }, + { + "epoch": 21.70846645367412, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 27179 + }, + { + "epoch": 21.70926517571885, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 27180 + }, + { + "epoch": 21.71006389776358, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27181 + }, + { + "epoch": 21.710862619808307, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 27182 + }, + { + "epoch": 21.711661341853034, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 27183 + }, + { + "epoch": 21.712460063897765, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 27184 + }, + { + "epoch": 21.713258785942493, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27185 + }, + { + "epoch": 21.71405750798722, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 27186 + }, + { + "epoch": 21.71485623003195, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 27187 + }, + { + "epoch": 21.715654952076676, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 27188 + }, + { + "epoch": 21.716453674121407, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 27189 + }, + { + "epoch": 21.717252396166135, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 27190 + }, + { + "epoch": 21.718051118210862, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 27191 + }, + { + "epoch": 21.71884984025559, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 27192 + }, + { + "epoch": 21.71964856230032, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 27193 + }, + { + "epoch": 21.72044728434505, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 27194 + }, + { + "epoch": 21.721246006389777, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 27195 + }, + { + "epoch": 21.722044728434504, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 27196 + }, + { + "epoch": 21.722843450479232, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 27197 + }, + { + "epoch": 21.723642172523963, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 27198 + }, + { + "epoch": 21.72444089456869, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 27199 + }, + { + "epoch": 21.72523961661342, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 27200 + }, + { + "epoch": 21.726038338658146, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 27201 + }, + { + "epoch": 21.726837060702877, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 27202 + }, + { + "epoch": 21.727635782747605, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 27203 + }, + { + "epoch": 21.728434504792332, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 27204 + }, + { + "epoch": 21.72923322683706, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 27205 + }, + { + "epoch": 21.730031948881788, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 27206 + }, + { + "epoch": 21.73083067092652, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 27207 + }, + { + "epoch": 21.731629392971247, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 27208 + }, + { + "epoch": 21.732428115015974, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 27209 + }, + { + "epoch": 21.733226837060702, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 27210 + }, + { + "epoch": 21.734025559105433, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27211 + }, + { + "epoch": 21.73482428115016, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27212 + }, + { + "epoch": 21.73562300319489, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 27213 + }, + { + "epoch": 21.736421725239616, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 27214 + }, + { + "epoch": 21.737220447284344, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 27215 + }, + { + "epoch": 21.738019169329075, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 27216 + }, + { + "epoch": 21.738817891373802, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 27217 + }, + { + "epoch": 21.73961661341853, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 27218 + }, + { + "epoch": 21.740415335463258, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 27219 + }, + { + "epoch": 21.74121405750799, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 27220 + }, + { + "epoch": 21.742012779552716, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 27221 + }, + { + "epoch": 21.742811501597444, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 27222 + }, + { + "epoch": 21.74361022364217, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27223 + }, + { + "epoch": 21.7444089456869, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 27224 + }, + { + "epoch": 21.74520766773163, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 27225 + }, + { + "epoch": 21.74600638977636, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 27226 + }, + { + "epoch": 21.746805111821086, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 27227 + }, + { + "epoch": 21.747603833865814, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 27228 + }, + { + "epoch": 21.748402555910545, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 27229 + }, + { + "epoch": 21.749201277955272, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 27230 + }, + { + "epoch": 21.75, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 27231 + }, + { + "epoch": 21.750798722044728, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27232 + }, + { + "epoch": 21.751597444089455, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 27233 + }, + { + "epoch": 21.752396166134186, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 27234 + }, + { + "epoch": 21.753194888178914, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 27235 + }, + { + "epoch": 21.75399361022364, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 27236 + }, + { + "epoch": 21.75479233226837, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 27237 + }, + { + "epoch": 21.7555910543131, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 27238 + }, + { + "epoch": 21.75638977635783, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 27239 + }, + { + "epoch": 21.757188498402556, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 27240 + }, + { + "epoch": 21.757987220447284, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 27241 + }, + { + "epoch": 21.75878594249201, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 27242 + }, + { + "epoch": 21.759584664536742, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 27243 + }, + { + "epoch": 21.76038338658147, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 27244 + }, + { + "epoch": 21.761182108626198, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 27245 + }, + { + "epoch": 21.761980830670925, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 27246 + }, + { + "epoch": 21.762779552715656, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 27247 + }, + { + "epoch": 21.763578274760384, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 27248 + }, + { + "epoch": 21.76437699680511, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 27249 + }, + { + "epoch": 21.76517571884984, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 27250 + }, + { + "epoch": 21.765974440894567, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 27251 + }, + { + "epoch": 21.766773162939298, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 27252 + }, + { + "epoch": 21.767571884984026, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 27253 + }, + { + "epoch": 21.768370607028753, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 27254 + }, + { + "epoch": 21.76916932907348, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 27255 + }, + { + "epoch": 21.769968051118212, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 27256 + }, + { + "epoch": 21.77076677316294, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 27257 + }, + { + "epoch": 21.771565495207668, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 27258 + }, + { + "epoch": 21.772364217252395, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 27259 + }, + { + "epoch": 21.773162939297123, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 27260 + }, + { + "epoch": 21.773961661341854, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27261 + }, + { + "epoch": 21.77476038338658, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 27262 + }, + { + "epoch": 21.77555910543131, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 27263 + }, + { + "epoch": 21.776357827476037, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 27264 + }, + { + "epoch": 21.777156549520768, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 27265 + }, + { + "epoch": 21.777955271565496, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 27266 + }, + { + "epoch": 21.778753993610223, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 27267 + }, + { + "epoch": 21.77955271565495, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 27268 + }, + { + "epoch": 21.78035143769968, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 27269 + }, + { + "epoch": 21.78115015974441, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 27270 + }, + { + "epoch": 21.781948881789138, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 27271 + }, + { + "epoch": 21.782747603833865, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 27272 + }, + { + "epoch": 21.783546325878593, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 27273 + }, + { + "epoch": 21.784345047923324, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 27274 + }, + { + "epoch": 21.78514376996805, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 27275 + }, + { + "epoch": 21.78594249201278, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 27276 + }, + { + "epoch": 21.786741214057507, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 27277 + }, + { + "epoch": 21.787539936102235, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27278 + }, + { + "epoch": 21.788338658146966, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 27279 + }, + { + "epoch": 21.789137380191693, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 27280 + }, + { + "epoch": 21.78993610223642, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 27281 + }, + { + "epoch": 21.79073482428115, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 27282 + }, + { + "epoch": 21.79153354632588, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27283 + }, + { + "epoch": 21.792332268370608, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 27284 + }, + { + "epoch": 21.793130990415335, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1065, + "step": 27285 + }, + { + "epoch": 21.793929712460063, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 27286 + }, + { + "epoch": 21.794728434504794, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 27287 + }, + { + "epoch": 21.79552715654952, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 27288 + }, + { + "epoch": 21.79632587859425, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 27289 + }, + { + "epoch": 21.797124600638977, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 27290 + }, + { + "epoch": 21.797923322683705, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 27291 + }, + { + "epoch": 21.798722044728436, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 27292 + }, + { + "epoch": 21.799520766773163, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 27293 + }, + { + "epoch": 21.80031948881789, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27294 + }, + { + "epoch": 21.80111821086262, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 27295 + }, + { + "epoch": 21.80191693290735, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 27296 + }, + { + "epoch": 21.802715654952078, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27297 + }, + { + "epoch": 21.803514376996805, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 27298 + }, + { + "epoch": 21.804313099041533, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 27299 + }, + { + "epoch": 21.80511182108626, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 27300 + }, + { + "epoch": 21.80591054313099, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 27301 + }, + { + "epoch": 21.80670926517572, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 27302 + }, + { + "epoch": 21.807507987220447, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27303 + }, + { + "epoch": 21.808306709265175, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 27304 + }, + { + "epoch": 21.809105431309906, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 27305 + }, + { + "epoch": 21.809904153354633, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 27306 + }, + { + "epoch": 21.81070287539936, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 27307 + }, + { + "epoch": 21.81150159744409, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 27308 + }, + { + "epoch": 21.812300319488816, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 27309 + }, + { + "epoch": 21.813099041533548, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 27310 + }, + { + "epoch": 21.813897763578275, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 27311 + }, + { + "epoch": 21.814696485623003, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 27312 + }, + { + "epoch": 21.81549520766773, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 27313 + }, + { + "epoch": 21.81629392971246, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 27314 + }, + { + "epoch": 21.81709265175719, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 27315 + }, + { + "epoch": 21.817891373801917, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 27316 + }, + { + "epoch": 21.818690095846645, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27317 + }, + { + "epoch": 21.819488817891372, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 27318 + }, + { + "epoch": 21.820287539936103, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 27319 + }, + { + "epoch": 21.82108626198083, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 27320 + }, + { + "epoch": 21.82188498402556, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27321 + }, + { + "epoch": 21.822683706070286, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 27322 + }, + { + "epoch": 21.823482428115017, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27323 + }, + { + "epoch": 21.824281150159745, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 27324 + }, + { + "epoch": 21.825079872204473, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 27325 + }, + { + "epoch": 21.8258785942492, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27326 + }, + { + "epoch": 21.826677316293928, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 27327 + }, + { + "epoch": 21.82747603833866, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27328 + }, + { + "epoch": 21.828274760383387, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 27329 + }, + { + "epoch": 21.829073482428115, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27330 + }, + { + "epoch": 21.829872204472842, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 27331 + }, + { + "epoch": 21.830670926517573, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 27332 + }, + { + "epoch": 21.8314696485623, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 27333 + }, + { + "epoch": 21.83226837060703, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 27334 + }, + { + "epoch": 21.833067092651756, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 27335 + }, + { + "epoch": 21.833865814696484, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 27336 + }, + { + "epoch": 21.834664536741215, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27337 + }, + { + "epoch": 21.835463258785943, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 27338 + }, + { + "epoch": 21.83626198083067, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 27339 + }, + { + "epoch": 21.837060702875398, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 27340 + }, + { + "epoch": 21.83785942492013, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 27341 + }, + { + "epoch": 21.838658146964857, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 27342 + }, + { + "epoch": 21.839456869009584, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 27343 + }, + { + "epoch": 21.840255591054312, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 27344 + }, + { + "epoch": 21.841054313099043, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 27345 + }, + { + "epoch": 21.84185303514377, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 27346 + }, + { + "epoch": 21.8426517571885, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1079, + "step": 27347 + }, + { + "epoch": 21.843450479233226, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 27348 + }, + { + "epoch": 21.844249201277954, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 27349 + }, + { + "epoch": 21.845047923322685, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 27350 + }, + { + "epoch": 21.845846645367413, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 27351 + }, + { + "epoch": 21.84664536741214, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 27352 + }, + { + "epoch": 21.847444089456868, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 27353 + }, + { + "epoch": 21.8482428115016, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 27354 + }, + { + "epoch": 21.849041533546327, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 27355 + }, + { + "epoch": 21.849840255591054, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 27356 + }, + { + "epoch": 21.850638977635782, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 27357 + }, + { + "epoch": 21.85143769968051, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 27358 + }, + { + "epoch": 21.85223642172524, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 27359 + }, + { + "epoch": 21.85303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 27360 + }, + { + "epoch": 21.853833865814696, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 27361 + }, + { + "epoch": 21.854632587859424, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 27362 + }, + { + "epoch": 21.855431309904155, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 27363 + }, + { + "epoch": 21.856230031948883, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 27364 + }, + { + "epoch": 21.85702875399361, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 27365 + }, + { + "epoch": 21.857827476038338, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 27366 + }, + { + "epoch": 21.858626198083066, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 27367 + }, + { + "epoch": 21.859424920127797, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 27368 + }, + { + "epoch": 21.860223642172524, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 27369 + }, + { + "epoch": 21.861022364217252, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 27370 + }, + { + "epoch": 21.86182108626198, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 27371 + }, + { + "epoch": 21.86261980830671, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 27372 + }, + { + "epoch": 21.86341853035144, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27373 + }, + { + "epoch": 21.864217252396166, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 27374 + }, + { + "epoch": 21.865015974440894, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 27375 + }, + { + "epoch": 21.86581469648562, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 27376 + }, + { + "epoch": 21.866613418530353, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 27377 + }, + { + "epoch": 21.86741214057508, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 27378 + }, + { + "epoch": 21.868210862619808, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 27379 + }, + { + "epoch": 21.869009584664536, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 27380 + }, + { + "epoch": 21.869808306709267, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 27381 + }, + { + "epoch": 21.870607028753994, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 27382 + }, + { + "epoch": 21.871405750798722, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 27383 + }, + { + "epoch": 21.87220447284345, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 27384 + }, + { + "epoch": 21.873003194888177, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27385 + }, + { + "epoch": 21.87380191693291, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 27386 + }, + { + "epoch": 21.874600638977636, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 27387 + }, + { + "epoch": 21.875399361022364, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 27388 + }, + { + "epoch": 21.87619808306709, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 27389 + }, + { + "epoch": 21.876996805111823, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 27390 + }, + { + "epoch": 21.87779552715655, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 27391 + }, + { + "epoch": 21.878594249201278, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 27392 + }, + { + "epoch": 21.879392971246006, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 27393 + }, + { + "epoch": 21.880191693290733, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 27394 + }, + { + "epoch": 21.880990415335464, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 27395 + }, + { + "epoch": 21.881789137380192, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 27396 + }, + { + "epoch": 21.88258785942492, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 27397 + }, + { + "epoch": 21.883386581469647, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27398 + }, + { + "epoch": 21.88418530351438, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 27399 + }, + { + "epoch": 21.884984025559106, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 27400 + }, + { + "epoch": 21.885782747603834, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 27401 + }, + { + "epoch": 21.88658146964856, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 27402 + }, + { + "epoch": 21.88738019169329, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 27403 + }, + { + "epoch": 21.88817891373802, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 27404 + }, + { + "epoch": 21.888977635782748, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 27405 + }, + { + "epoch": 21.889776357827476, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 27406 + }, + { + "epoch": 21.890575079872203, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 27407 + }, + { + "epoch": 21.891373801916934, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 27408 + }, + { + "epoch": 21.892172523961662, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 27409 + }, + { + "epoch": 21.89297124600639, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 27410 + }, + { + "epoch": 21.893769968051117, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 27411 + }, + { + "epoch": 21.894568690095845, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27412 + }, + { + "epoch": 21.895367412140576, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 27413 + }, + { + "epoch": 21.896166134185304, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 27414 + }, + { + "epoch": 21.89696485623003, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 27415 + }, + { + "epoch": 21.89776357827476, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 27416 + }, + { + "epoch": 21.89856230031949, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 27417 + }, + { + "epoch": 21.899361022364218, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 27418 + }, + { + "epoch": 21.900159744408946, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 27419 + }, + { + "epoch": 21.900958466453673, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 27420 + }, + { + "epoch": 21.9017571884984, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 27421 + }, + { + "epoch": 21.902555910543132, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 27422 + }, + { + "epoch": 21.90335463258786, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 27423 + }, + { + "epoch": 21.904153354632587, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 27424 + }, + { + "epoch": 21.904952076677315, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 27425 + }, + { + "epoch": 21.905750798722046, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 27426 + }, + { + "epoch": 21.906549520766774, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27427 + }, + { + "epoch": 21.9073482428115, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 27428 + }, + { + "epoch": 21.90814696485623, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 27429 + }, + { + "epoch": 21.908945686900957, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 27430 + }, + { + "epoch": 21.909744408945688, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 27431 + }, + { + "epoch": 21.910543130990416, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 27432 + }, + { + "epoch": 21.911341853035143, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27433 + }, + { + "epoch": 21.91214057507987, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 27434 + }, + { + "epoch": 21.912939297124602, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 27435 + }, + { + "epoch": 21.91373801916933, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27436 + }, + { + "epoch": 21.914536741214057, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 27437 + }, + { + "epoch": 21.915335463258785, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 27438 + }, + { + "epoch": 21.916134185303516, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 27439 + }, + { + "epoch": 21.916932907348244, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 27440 + }, + { + "epoch": 21.91773162939297, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 27441 + }, + { + "epoch": 21.9185303514377, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 27442 + }, + { + "epoch": 21.919329073482427, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 27443 + }, + { + "epoch": 21.920127795527158, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 27444 + }, + { + "epoch": 21.920926517571885, + "grad_norm": 0.3515625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 27445 + }, + { + "epoch": 21.921725239616613, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 27446 + }, + { + "epoch": 21.92252396166134, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 27447 + }, + { + "epoch": 21.923322683706072, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 27448 + }, + { + "epoch": 21.9241214057508, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 27449 + }, + { + "epoch": 21.924920127795527, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 27450 + }, + { + "epoch": 21.925718849840255, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 27451 + }, + { + "epoch": 21.926517571884983, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 27452 + }, + { + "epoch": 21.927316293929714, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 27453 + }, + { + "epoch": 21.92811501597444, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 27454 + }, + { + "epoch": 21.92891373801917, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 27455 + }, + { + "epoch": 21.929712460063897, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27456 + }, + { + "epoch": 21.930511182108628, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 27457 + }, + { + "epoch": 21.931309904153355, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 27458 + }, + { + "epoch": 21.932108626198083, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 27459 + }, + { + "epoch": 21.93290734824281, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 27460 + }, + { + "epoch": 21.93370607028754, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 27461 + }, + { + "epoch": 21.93450479233227, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 27462 + }, + { + "epoch": 21.935303514376997, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 27463 + }, + { + "epoch": 21.936102236421725, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 27464 + }, + { + "epoch": 21.936900958466452, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 27465 + }, + { + "epoch": 21.937699680511184, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 27466 + }, + { + "epoch": 21.93849840255591, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 27467 + }, + { + "epoch": 21.93929712460064, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 27468 + }, + { + "epoch": 21.940095846645367, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 27469 + }, + { + "epoch": 21.940894568690094, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27470 + }, + { + "epoch": 21.941693290734825, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 27471 + }, + { + "epoch": 21.942492012779553, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 27472 + }, + { + "epoch": 21.94329073482428, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 27473 + }, + { + "epoch": 21.94408945686901, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1049, + "step": 27474 + }, + { + "epoch": 21.94488817891374, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 27475 + }, + { + "epoch": 21.945686900958467, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 27476 + }, + { + "epoch": 21.946485623003195, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 27477 + }, + { + "epoch": 21.947284345047922, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 27478 + }, + { + "epoch": 21.94808306709265, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 27479 + }, + { + "epoch": 21.94888178913738, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 27480 + }, + { + "epoch": 21.94968051118211, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 27481 + }, + { + "epoch": 21.950479233226837, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 27482 + }, + { + "epoch": 21.951277955271564, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 27483 + }, + { + "epoch": 21.952076677316295, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 27484 + }, + { + "epoch": 21.952875399361023, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 27485 + }, + { + "epoch": 21.95367412140575, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 27486 + }, + { + "epoch": 21.95447284345048, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 27487 + }, + { + "epoch": 21.955271565495206, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 27488 + }, + { + "epoch": 21.956070287539937, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27489 + }, + { + "epoch": 21.956869009584665, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 27490 + }, + { + "epoch": 21.957667731629392, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 27491 + }, + { + "epoch": 21.95846645367412, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 27492 + }, + { + "epoch": 21.95926517571885, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27493 + }, + { + "epoch": 21.96006389776358, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 27494 + }, + { + "epoch": 21.960862619808307, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 27495 + }, + { + "epoch": 21.961661341853034, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 27496 + }, + { + "epoch": 21.962460063897765, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 27497 + }, + { + "epoch": 21.963258785942493, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 27498 + }, + { + "epoch": 21.96405750798722, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 27499 + }, + { + "epoch": 21.96485623003195, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 27500 + }, + { + "epoch": 21.965654952076676, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 27501 + }, + { + "epoch": 21.966453674121407, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 27502 + }, + { + "epoch": 21.967252396166135, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 27503 + }, + { + "epoch": 21.968051118210862, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 27504 + }, + { + "epoch": 21.96884984025559, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27505 + }, + { + "epoch": 21.96964856230032, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27506 + }, + { + "epoch": 21.97044728434505, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 27507 + }, + { + "epoch": 21.971246006389777, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27508 + }, + { + "epoch": 21.972044728434504, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27509 + }, + { + "epoch": 21.972843450479232, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 27510 + }, + { + "epoch": 21.973642172523963, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 27511 + }, + { + "epoch": 21.97444089456869, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 27512 + }, + { + "epoch": 21.97523961661342, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 27513 + }, + { + "epoch": 21.976038338658146, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27514 + }, + { + "epoch": 21.976837060702877, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 27515 + }, + { + "epoch": 21.977635782747605, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 27516 + }, + { + "epoch": 21.978434504792332, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27517 + }, + { + "epoch": 21.97923322683706, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 27518 + }, + { + "epoch": 21.980031948881788, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 27519 + }, + { + "epoch": 21.98083067092652, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27520 + }, + { + "epoch": 21.981629392971247, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 27521 + }, + { + "epoch": 21.982428115015974, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 27522 + }, + { + "epoch": 21.983226837060702, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 27523 + }, + { + "epoch": 21.984025559105433, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0725, + "step": 27524 + }, + { + "epoch": 21.98482428115016, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 27525 + }, + { + "epoch": 21.98562300319489, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27526 + }, + { + "epoch": 21.986421725239616, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 27527 + }, + { + "epoch": 21.987220447284344, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 27528 + }, + { + "epoch": 21.988019169329075, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27529 + }, + { + "epoch": 21.988817891373802, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27530 + }, + { + "epoch": 21.98961661341853, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 27531 + }, + { + "epoch": 21.990415335463258, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 27532 + }, + { + "epoch": 21.99121405750799, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 27533 + }, + { + "epoch": 21.992012779552716, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 27534 + }, + { + "epoch": 21.992811501597444, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 27535 + }, + { + "epoch": 21.99361022364217, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 27536 + }, + { + "epoch": 21.9944089456869, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27537 + }, + { + "epoch": 21.99520766773163, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 27538 + }, + { + "epoch": 21.99600638977636, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 27539 + }, + { + "epoch": 21.996805111821086, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 27540 + }, + { + "epoch": 21.997603833865814, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27541 + }, + { + "epoch": 21.998402555910545, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 27542 + }, + { + "epoch": 21.999201277955272, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 27543 + }, + { + "epoch": 22.0, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 27544 + }, + { + "epoch": 22.000798722044728, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 27545 + }, + { + "epoch": 22.001597444089455, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 27546 + }, + { + "epoch": 22.002396166134186, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 27547 + }, + { + "epoch": 22.003194888178914, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 27548 + }, + { + "epoch": 22.00399361022364, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 27549 + }, + { + "epoch": 22.00479233226837, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 27550 + }, + { + "epoch": 22.0055910543131, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 27551 + }, + { + "epoch": 22.00638977635783, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 27552 + }, + { + "epoch": 22.007188498402556, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 27553 + }, + { + "epoch": 22.007987220447284, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 27554 + }, + { + "epoch": 22.00878594249201, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 27555 + }, + { + "epoch": 22.009584664536742, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 27556 + }, + { + "epoch": 22.01038338658147, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 27557 + }, + { + "epoch": 22.011182108626198, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 27558 + }, + { + "epoch": 22.011980830670925, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 27559 + }, + { + "epoch": 22.012779552715656, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 27560 + }, + { + "epoch": 22.013578274760384, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 27561 + }, + { + "epoch": 22.01437699680511, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 27562 + }, + { + "epoch": 22.01517571884984, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 27563 + }, + { + "epoch": 22.015974440894567, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 27564 + }, + { + "epoch": 22.016773162939298, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 27565 + }, + { + "epoch": 22.017571884984026, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 27566 + }, + { + "epoch": 22.018370607028753, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 27567 + }, + { + "epoch": 22.01916932907348, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 27568 + }, + { + "epoch": 22.019968051118212, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 27569 + }, + { + "epoch": 22.02076677316294, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27570 + }, + { + "epoch": 22.021565495207668, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 27571 + }, + { + "epoch": 22.022364217252395, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 27572 + }, + { + "epoch": 22.023162939297123, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27573 + }, + { + "epoch": 22.023961661341854, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 27574 + }, + { + "epoch": 22.02476038338658, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 27575 + }, + { + "epoch": 22.02555910543131, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 27576 + }, + { + "epoch": 22.026357827476037, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 27577 + }, + { + "epoch": 22.027156549520768, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 27578 + }, + { + "epoch": 22.027955271565496, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 27579 + }, + { + "epoch": 22.028753993610223, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27580 + }, + { + "epoch": 22.02955271565495, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 27581 + }, + { + "epoch": 22.03035143769968, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27582 + }, + { + "epoch": 22.03115015974441, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 27583 + }, + { + "epoch": 22.031948881789138, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 27584 + }, + { + "epoch": 22.032747603833865, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27585 + }, + { + "epoch": 22.033546325878593, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 27586 + }, + { + "epoch": 22.034345047923324, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 27587 + }, + { + "epoch": 22.03514376996805, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 27588 + }, + { + "epoch": 22.03594249201278, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 27589 + }, + { + "epoch": 22.036741214057507, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 27590 + }, + { + "epoch": 22.037539936102238, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 27591 + }, + { + "epoch": 22.038338658146966, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 27592 + }, + { + "epoch": 22.039137380191693, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 27593 + }, + { + "epoch": 22.03993610223642, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 27594 + }, + { + "epoch": 22.04073482428115, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27595 + }, + { + "epoch": 22.04153354632588, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 27596 + }, + { + "epoch": 22.042332268370608, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 27597 + }, + { + "epoch": 22.043130990415335, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 27598 + }, + { + "epoch": 22.043929712460063, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 27599 + }, + { + "epoch": 22.044728434504794, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 27600 + }, + { + "epoch": 22.04552715654952, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 27601 + }, + { + "epoch": 22.04632587859425, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 27602 + }, + { + "epoch": 22.047124600638977, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 27603 + }, + { + "epoch": 22.047923322683705, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 27604 + }, + { + "epoch": 22.048722044728436, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 27605 + }, + { + "epoch": 22.049520766773163, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 27606 + }, + { + "epoch": 22.05031948881789, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 27607 + }, + { + "epoch": 22.05111821086262, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 27608 + }, + { + "epoch": 22.05191693290735, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 27609 + }, + { + "epoch": 22.052715654952078, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 27610 + }, + { + "epoch": 22.053514376996805, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 27611 + }, + { + "epoch": 22.054313099041533, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27612 + }, + { + "epoch": 22.05511182108626, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 27613 + }, + { + "epoch": 22.05591054313099, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 27614 + }, + { + "epoch": 22.05670926517572, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 27615 + }, + { + "epoch": 22.057507987220447, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 27616 + }, + { + "epoch": 22.058306709265175, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 27617 + }, + { + "epoch": 22.059105431309906, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 27618 + }, + { + "epoch": 22.059904153354633, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 27619 + }, + { + "epoch": 22.06070287539936, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 27620 + }, + { + "epoch": 22.06150159744409, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27621 + }, + { + "epoch": 22.062300319488816, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 27622 + }, + { + "epoch": 22.063099041533548, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27623 + }, + { + "epoch": 22.063897763578275, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 27624 + }, + { + "epoch": 22.064696485623003, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 27625 + }, + { + "epoch": 22.06549520766773, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 27626 + }, + { + "epoch": 22.06629392971246, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 27627 + }, + { + "epoch": 22.06709265175719, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 27628 + }, + { + "epoch": 22.067891373801917, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 27629 + }, + { + "epoch": 22.068690095846645, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 27630 + }, + { + "epoch": 22.069488817891372, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 27631 + }, + { + "epoch": 22.070287539936103, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 27632 + }, + { + "epoch": 22.07108626198083, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 27633 + }, + { + "epoch": 22.07188498402556, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 27634 + }, + { + "epoch": 22.072683706070286, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 27635 + }, + { + "epoch": 22.073482428115017, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 27636 + }, + { + "epoch": 22.074281150159745, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 27637 + }, + { + "epoch": 22.075079872204473, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 27638 + }, + { + "epoch": 22.0758785942492, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27639 + }, + { + "epoch": 22.076677316293928, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 27640 + }, + { + "epoch": 22.07747603833866, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 27641 + }, + { + "epoch": 22.078274760383387, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 27642 + }, + { + "epoch": 22.079073482428115, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 27643 + }, + { + "epoch": 22.079872204472842, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 27644 + }, + { + "epoch": 22.080670926517573, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 27645 + }, + { + "epoch": 22.0814696485623, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 27646 + }, + { + "epoch": 22.08226837060703, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 27647 + }, + { + "epoch": 22.083067092651756, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 27648 + }, + { + "epoch": 22.083865814696484, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 27649 + }, + { + "epoch": 22.084664536741215, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 27650 + }, + { + "epoch": 22.085463258785943, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 27651 + }, + { + "epoch": 22.08626198083067, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 27652 + }, + { + "epoch": 22.087060702875398, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 27653 + }, + { + "epoch": 22.08785942492013, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 27654 + }, + { + "epoch": 22.088658146964857, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 27655 + }, + { + "epoch": 22.089456869009584, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 27656 + }, + { + "epoch": 22.090255591054312, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 27657 + }, + { + "epoch": 22.09105431309904, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 27658 + }, + { + "epoch": 22.09185303514377, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 27659 + }, + { + "epoch": 22.0926517571885, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 27660 + }, + { + "epoch": 22.093450479233226, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 27661 + }, + { + "epoch": 22.094249201277954, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27662 + }, + { + "epoch": 22.095047923322685, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27663 + }, + { + "epoch": 22.095846645367413, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 27664 + }, + { + "epoch": 22.09664536741214, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 27665 + }, + { + "epoch": 22.097444089456868, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 27666 + }, + { + "epoch": 22.0982428115016, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 27667 + }, + { + "epoch": 22.099041533546327, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 27668 + }, + { + "epoch": 22.099840255591054, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 27669 + }, + { + "epoch": 22.100638977635782, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27670 + }, + { + "epoch": 22.10143769968051, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 27671 + }, + { + "epoch": 22.10223642172524, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 27672 + }, + { + "epoch": 22.10303514376997, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 27673 + }, + { + "epoch": 22.103833865814696, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 27674 + }, + { + "epoch": 22.104632587859424, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 27675 + }, + { + "epoch": 22.105431309904155, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 27676 + }, + { + "epoch": 22.106230031948883, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27677 + }, + { + "epoch": 22.10702875399361, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 27678 + }, + { + "epoch": 22.107827476038338, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 27679 + }, + { + "epoch": 22.108626198083066, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27680 + }, + { + "epoch": 22.109424920127797, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 27681 + }, + { + "epoch": 22.110223642172524, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 27682 + }, + { + "epoch": 22.111022364217252, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 27683 + }, + { + "epoch": 22.11182108626198, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 27684 + }, + { + "epoch": 22.11261980830671, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 27685 + }, + { + "epoch": 22.11341853035144, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 27686 + }, + { + "epoch": 22.114217252396166, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 27687 + }, + { + "epoch": 22.115015974440894, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 27688 + }, + { + "epoch": 22.11581469648562, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 27689 + }, + { + "epoch": 22.116613418530353, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 27690 + }, + { + "epoch": 22.11741214057508, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 27691 + }, + { + "epoch": 22.118210862619808, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 27692 + }, + { + "epoch": 22.119009584664536, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 27693 + }, + { + "epoch": 22.119808306709267, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 27694 + }, + { + "epoch": 22.120607028753994, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 27695 + }, + { + "epoch": 22.121405750798722, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 27696 + }, + { + "epoch": 22.12220447284345, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 27697 + }, + { + "epoch": 22.123003194888177, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 27698 + }, + { + "epoch": 22.12380191693291, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 27699 + }, + { + "epoch": 22.124600638977636, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 27700 + }, + { + "epoch": 22.125399361022364, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 27701 + }, + { + "epoch": 22.12619808306709, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 27702 + }, + { + "epoch": 22.126996805111823, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27703 + }, + { + "epoch": 22.12779552715655, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 27704 + }, + { + "epoch": 22.128594249201278, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 27705 + }, + { + "epoch": 22.129392971246006, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 27706 + }, + { + "epoch": 22.130191693290733, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 27707 + }, + { + "epoch": 22.130990415335464, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27708 + }, + { + "epoch": 22.131789137380192, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 27709 + }, + { + "epoch": 22.13258785942492, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 27710 + }, + { + "epoch": 22.133386581469647, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 27711 + }, + { + "epoch": 22.13418530351438, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 27712 + }, + { + "epoch": 22.134984025559106, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 27713 + }, + { + "epoch": 22.135782747603834, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 27714 + }, + { + "epoch": 22.13658146964856, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 27715 + }, + { + "epoch": 22.13738019169329, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 27716 + }, + { + "epoch": 22.13817891373802, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 27717 + }, + { + "epoch": 22.138977635782748, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 27718 + }, + { + "epoch": 22.139776357827476, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 27719 + }, + { + "epoch": 22.140575079872203, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 27720 + }, + { + "epoch": 22.141373801916934, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 27721 + }, + { + "epoch": 22.142172523961662, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 27722 + }, + { + "epoch": 22.14297124600639, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 27723 + }, + { + "epoch": 22.143769968051117, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 27724 + }, + { + "epoch": 22.144568690095845, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 27725 + }, + { + "epoch": 22.145367412140576, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 27726 + }, + { + "epoch": 22.146166134185304, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 27727 + }, + { + "epoch": 22.14696485623003, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 27728 + }, + { + "epoch": 22.14776357827476, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 27729 + }, + { + "epoch": 22.14856230031949, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 27730 + }, + { + "epoch": 22.149361022364218, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 27731 + }, + { + "epoch": 22.150159744408946, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 27732 + }, + { + "epoch": 22.150958466453673, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 27733 + }, + { + "epoch": 22.1517571884984, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 27734 + }, + { + "epoch": 22.152555910543132, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 27735 + }, + { + "epoch": 22.15335463258786, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 27736 + }, + { + "epoch": 22.154153354632587, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 27737 + }, + { + "epoch": 22.154952076677315, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 27738 + }, + { + "epoch": 22.155750798722046, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 27739 + }, + { + "epoch": 22.156549520766774, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 27740 + }, + { + "epoch": 22.1573482428115, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 27741 + }, + { + "epoch": 22.15814696485623, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 27742 + }, + { + "epoch": 22.15894568690096, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 27743 + }, + { + "epoch": 22.159744408945688, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 27744 + }, + { + "epoch": 22.160543130990416, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 27745 + }, + { + "epoch": 22.161341853035143, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 27746 + }, + { + "epoch": 22.16214057507987, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 27747 + }, + { + "epoch": 22.162939297124602, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 27748 + }, + { + "epoch": 22.16373801916933, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 27749 + }, + { + "epoch": 22.164536741214057, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 27750 + }, + { + "epoch": 22.165335463258785, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 27751 + }, + { + "epoch": 22.166134185303516, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 27752 + }, + { + "epoch": 22.166932907348244, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 27753 + }, + { + "epoch": 22.16773162939297, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 27754 + }, + { + "epoch": 22.1685303514377, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 27755 + }, + { + "epoch": 22.169329073482427, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 27756 + }, + { + "epoch": 22.170127795527158, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 27757 + }, + { + "epoch": 22.170926517571885, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 27758 + }, + { + "epoch": 22.171725239616613, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 27759 + }, + { + "epoch": 22.17252396166134, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 27760 + }, + { + "epoch": 22.173322683706072, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 27761 + }, + { + "epoch": 22.1741214057508, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 27762 + }, + { + "epoch": 22.174920127795527, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27763 + }, + { + "epoch": 22.175718849840255, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 27764 + }, + { + "epoch": 22.176517571884983, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 27765 + }, + { + "epoch": 22.177316293929714, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 27766 + }, + { + "epoch": 22.17811501597444, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 27767 + }, + { + "epoch": 22.17891373801917, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 27768 + }, + { + "epoch": 22.179712460063897, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27769 + }, + { + "epoch": 22.180511182108628, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 27770 + }, + { + "epoch": 22.181309904153355, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 27771 + }, + { + "epoch": 22.182108626198083, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27772 + }, + { + "epoch": 22.18290734824281, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1031, + "step": 27773 + }, + { + "epoch": 22.18370607028754, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 27774 + }, + { + "epoch": 22.18450479233227, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 27775 + }, + { + "epoch": 22.185303514376997, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 27776 + }, + { + "epoch": 22.186102236421725, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 27777 + }, + { + "epoch": 22.186900958466452, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 27778 + }, + { + "epoch": 22.187699680511184, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 27779 + }, + { + "epoch": 22.18849840255591, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 27780 + }, + { + "epoch": 22.18929712460064, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 27781 + }, + { + "epoch": 22.190095846645367, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 27782 + }, + { + "epoch": 22.190894568690094, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 27783 + }, + { + "epoch": 22.191693290734825, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 27784 + }, + { + "epoch": 22.192492012779553, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 27785 + }, + { + "epoch": 22.19329073482428, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 27786 + }, + { + "epoch": 22.19408945686901, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 27787 + }, + { + "epoch": 22.19488817891374, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27788 + }, + { + "epoch": 22.195686900958467, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 27789 + }, + { + "epoch": 22.196485623003195, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 27790 + }, + { + "epoch": 22.197284345047922, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 27791 + }, + { + "epoch": 22.19808306709265, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 27792 + }, + { + "epoch": 22.19888178913738, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 27793 + }, + { + "epoch": 22.19968051118211, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 27794 + }, + { + "epoch": 22.200479233226837, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 27795 + }, + { + "epoch": 22.201277955271564, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 27796 + }, + { + "epoch": 22.202076677316295, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 27797 + }, + { + "epoch": 22.202875399361023, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 27798 + }, + { + "epoch": 22.20367412140575, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 27799 + }, + { + "epoch": 22.20447284345048, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 27800 + }, + { + "epoch": 22.205271565495206, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 27801 + }, + { + "epoch": 22.206070287539937, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 27802 + }, + { + "epoch": 22.206869009584665, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 27803 + }, + { + "epoch": 22.207667731629392, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 27804 + }, + { + "epoch": 22.20846645367412, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 27805 + }, + { + "epoch": 22.20926517571885, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 27806 + }, + { + "epoch": 22.21006389776358, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 27807 + }, + { + "epoch": 22.210862619808307, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 27808 + }, + { + "epoch": 22.211661341853034, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 27809 + }, + { + "epoch": 22.212460063897762, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 27810 + }, + { + "epoch": 22.213258785942493, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27811 + }, + { + "epoch": 22.21405750798722, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 27812 + }, + { + "epoch": 22.21485623003195, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 27813 + }, + { + "epoch": 22.215654952076676, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 27814 + }, + { + "epoch": 22.216453674121407, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 27815 + }, + { + "epoch": 22.217252396166135, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 27816 + }, + { + "epoch": 22.218051118210862, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27817 + }, + { + "epoch": 22.21884984025559, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 27818 + }, + { + "epoch": 22.21964856230032, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 27819 + }, + { + "epoch": 22.22044728434505, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 27820 + }, + { + "epoch": 22.221246006389777, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 27821 + }, + { + "epoch": 22.222044728434504, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 27822 + }, + { + "epoch": 22.222843450479232, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 27823 + }, + { + "epoch": 22.223642172523963, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 27824 + }, + { + "epoch": 22.22444089456869, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 27825 + }, + { + "epoch": 22.22523961661342, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 27826 + }, + { + "epoch": 22.226038338658146, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 27827 + }, + { + "epoch": 22.226837060702877, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 27828 + }, + { + "epoch": 22.227635782747605, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 27829 + }, + { + "epoch": 22.228434504792332, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 27830 + }, + { + "epoch": 22.22923322683706, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 27831 + }, + { + "epoch": 22.230031948881788, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 27832 + }, + { + "epoch": 22.23083067092652, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 27833 + }, + { + "epoch": 22.231629392971247, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 27834 + }, + { + "epoch": 22.232428115015974, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 27835 + }, + { + "epoch": 22.233226837060702, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 27836 + }, + { + "epoch": 22.234025559105433, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 27837 + }, + { + "epoch": 22.23482428115016, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 27838 + }, + { + "epoch": 22.23562300319489, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 27839 + }, + { + "epoch": 22.236421725239616, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 27840 + }, + { + "epoch": 22.237220447284344, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 27841 + }, + { + "epoch": 22.238019169329075, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 27842 + }, + { + "epoch": 22.238817891373802, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 27843 + }, + { + "epoch": 22.23961661341853, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 27844 + }, + { + "epoch": 22.240415335463258, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 27845 + }, + { + "epoch": 22.24121405750799, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 27846 + }, + { + "epoch": 22.242012779552716, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 27847 + }, + { + "epoch": 22.242811501597444, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 27848 + }, + { + "epoch": 22.24361022364217, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 27849 + }, + { + "epoch": 22.2444089456869, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 27850 + }, + { + "epoch": 22.24520766773163, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 27851 + }, + { + "epoch": 22.24600638977636, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 27852 + }, + { + "epoch": 22.246805111821086, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 27853 + }, + { + "epoch": 22.247603833865814, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 27854 + }, + { + "epoch": 22.248402555910545, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 27855 + }, + { + "epoch": 22.249201277955272, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 27856 + }, + { + "epoch": 22.25, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 27857 + }, + { + "epoch": 22.250798722044728, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 27858 + }, + { + "epoch": 22.251597444089455, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 27859 + }, + { + "epoch": 22.252396166134186, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 27860 + }, + { + "epoch": 22.253194888178914, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 27861 + }, + { + "epoch": 22.25399361022364, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 27862 + }, + { + "epoch": 22.25479233226837, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 27863 + }, + { + "epoch": 22.2555910543131, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 27864 + }, + { + "epoch": 22.25638977635783, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 27865 + }, + { + "epoch": 22.257188498402556, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 27866 + }, + { + "epoch": 22.257987220447284, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 27867 + }, + { + "epoch": 22.25878594249201, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 27868 + }, + { + "epoch": 22.259584664536742, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 27869 + }, + { + "epoch": 22.26038338658147, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 27870 + }, + { + "epoch": 22.261182108626198, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 27871 + }, + { + "epoch": 22.261980830670925, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 27872 + }, + { + "epoch": 22.262779552715656, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 27873 + }, + { + "epoch": 22.263578274760384, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 27874 + }, + { + "epoch": 22.26437699680511, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 27875 + }, + { + "epoch": 22.26517571884984, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 27876 + }, + { + "epoch": 22.265974440894567, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 27877 + }, + { + "epoch": 22.266773162939298, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 27878 + }, + { + "epoch": 22.267571884984026, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 27879 + }, + { + "epoch": 22.268370607028753, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 27880 + }, + { + "epoch": 22.26916932907348, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 27881 + }, + { + "epoch": 22.269968051118212, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 27882 + }, + { + "epoch": 22.27076677316294, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 27883 + }, + { + "epoch": 22.271565495207668, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 27884 + }, + { + "epoch": 22.272364217252395, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 27885 + }, + { + "epoch": 22.273162939297123, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27886 + }, + { + "epoch": 22.273961661341854, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 27887 + }, + { + "epoch": 22.27476038338658, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 27888 + }, + { + "epoch": 22.27555910543131, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 27889 + }, + { + "epoch": 22.276357827476037, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27890 + }, + { + "epoch": 22.277156549520768, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 27891 + }, + { + "epoch": 22.277955271565496, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 27892 + }, + { + "epoch": 22.278753993610223, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 27893 + }, + { + "epoch": 22.27955271565495, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 27894 + }, + { + "epoch": 22.28035143769968, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 27895 + }, + { + "epoch": 22.28115015974441, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 27896 + }, + { + "epoch": 22.281948881789138, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 27897 + }, + { + "epoch": 22.282747603833865, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 27898 + }, + { + "epoch": 22.283546325878593, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 27899 + }, + { + "epoch": 22.284345047923324, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 27900 + }, + { + "epoch": 22.28514376996805, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 27901 + }, + { + "epoch": 22.28594249201278, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 27902 + }, + { + "epoch": 22.286741214057507, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27903 + }, + { + "epoch": 22.287539936102238, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 27904 + }, + { + "epoch": 22.288338658146966, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 27905 + }, + { + "epoch": 22.289137380191693, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 27906 + }, + { + "epoch": 22.28993610223642, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 27907 + }, + { + "epoch": 22.29073482428115, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 27908 + }, + { + "epoch": 22.29153354632588, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 27909 + }, + { + "epoch": 22.292332268370608, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 27910 + }, + { + "epoch": 22.293130990415335, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 27911 + }, + { + "epoch": 22.293929712460063, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 27912 + }, + { + "epoch": 22.294728434504794, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 27913 + }, + { + "epoch": 22.29552715654952, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 27914 + }, + { + "epoch": 22.29632587859425, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 27915 + }, + { + "epoch": 22.297124600638977, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 27916 + }, + { + "epoch": 22.297923322683705, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 27917 + }, + { + "epoch": 22.298722044728436, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 27918 + }, + { + "epoch": 22.299520766773163, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 27919 + }, + { + "epoch": 22.30031948881789, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 27920 + }, + { + "epoch": 22.30111821086262, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 27921 + }, + { + "epoch": 22.30191693290735, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 27922 + }, + { + "epoch": 22.302715654952078, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 27923 + }, + { + "epoch": 22.303514376996805, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 27924 + }, + { + "epoch": 22.304313099041533, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 27925 + }, + { + "epoch": 22.30511182108626, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 27926 + }, + { + "epoch": 22.30591054313099, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 27927 + }, + { + "epoch": 22.30670926517572, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 27928 + }, + { + "epoch": 22.307507987220447, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 27929 + }, + { + "epoch": 22.308306709265175, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 27930 + }, + { + "epoch": 22.309105431309906, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 27931 + }, + { + "epoch": 22.309904153354633, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 27932 + }, + { + "epoch": 22.31070287539936, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 27933 + }, + { + "epoch": 22.31150159744409, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 27934 + }, + { + "epoch": 22.312300319488816, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 27935 + }, + { + "epoch": 22.313099041533548, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 27936 + }, + { + "epoch": 22.313897763578275, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 27937 + }, + { + "epoch": 22.314696485623003, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 27938 + }, + { + "epoch": 22.31549520766773, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 27939 + }, + { + "epoch": 22.31629392971246, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 27940 + }, + { + "epoch": 22.31709265175719, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 27941 + }, + { + "epoch": 22.317891373801917, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 27942 + }, + { + "epoch": 22.318690095846645, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 27943 + }, + { + "epoch": 22.319488817891372, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 27944 + }, + { + "epoch": 22.320287539936103, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 27945 + }, + { + "epoch": 22.32108626198083, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 27946 + }, + { + "epoch": 22.32188498402556, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 27947 + }, + { + "epoch": 22.322683706070286, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 27948 + }, + { + "epoch": 22.323482428115017, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 27949 + }, + { + "epoch": 22.324281150159745, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 27950 + }, + { + "epoch": 22.325079872204473, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 27951 + }, + { + "epoch": 22.3258785942492, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 27952 + }, + { + "epoch": 22.326677316293928, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 27953 + }, + { + "epoch": 22.32747603833866, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 27954 + }, + { + "epoch": 22.328274760383387, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 27955 + }, + { + "epoch": 22.329073482428115, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 27956 + }, + { + "epoch": 22.329872204472842, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 27957 + }, + { + "epoch": 22.330670926517573, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 27958 + }, + { + "epoch": 22.3314696485623, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 27959 + }, + { + "epoch": 22.33226837060703, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 27960 + }, + { + "epoch": 22.333067092651756, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 27961 + }, + { + "epoch": 22.333865814696484, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 27962 + }, + { + "epoch": 22.334664536741215, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 27963 + }, + { + "epoch": 22.335463258785943, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 27964 + }, + { + "epoch": 22.33626198083067, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 27965 + }, + { + "epoch": 22.337060702875398, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 27966 + }, + { + "epoch": 22.33785942492013, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 27967 + }, + { + "epoch": 22.338658146964857, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 27968 + }, + { + "epoch": 22.339456869009584, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 27969 + }, + { + "epoch": 22.340255591054312, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 27970 + }, + { + "epoch": 22.341054313099043, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 27971 + }, + { + "epoch": 22.34185303514377, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 27972 + }, + { + "epoch": 22.3426517571885, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 27973 + }, + { + "epoch": 22.343450479233226, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27974 + }, + { + "epoch": 22.344249201277954, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 27975 + }, + { + "epoch": 22.345047923322685, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 27976 + }, + { + "epoch": 22.345846645367413, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 27977 + }, + { + "epoch": 22.34664536741214, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 27978 + }, + { + "epoch": 22.347444089456868, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 27979 + }, + { + "epoch": 22.3482428115016, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 27980 + }, + { + "epoch": 22.349041533546327, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 27981 + }, + { + "epoch": 22.349840255591054, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 27982 + }, + { + "epoch": 22.350638977635782, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 27983 + }, + { + "epoch": 22.35143769968051, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 27984 + }, + { + "epoch": 22.35223642172524, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 27985 + }, + { + "epoch": 22.35303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 27986 + }, + { + "epoch": 22.353833865814696, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 27987 + }, + { + "epoch": 22.354632587859424, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 27988 + }, + { + "epoch": 22.355431309904155, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 27989 + }, + { + "epoch": 22.356230031948883, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 27990 + }, + { + "epoch": 22.35702875399361, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 27991 + }, + { + "epoch": 22.357827476038338, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 27992 + }, + { + "epoch": 22.358626198083066, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 27993 + }, + { + "epoch": 22.359424920127797, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 27994 + }, + { + "epoch": 22.360223642172524, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 27995 + }, + { + "epoch": 22.361022364217252, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 27996 + }, + { + "epoch": 22.36182108626198, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 27997 + }, + { + "epoch": 22.36261980830671, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 27998 + }, + { + "epoch": 22.36341853035144, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 27999 + }, + { + "epoch": 22.364217252396166, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 28000 + }, + { + "epoch": 22.365015974440894, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 28001 + }, + { + "epoch": 22.36581469648562, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 28002 + }, + { + "epoch": 22.366613418530353, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 28003 + }, + { + "epoch": 22.36741214057508, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 28004 + }, + { + "epoch": 22.368210862619808, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 28005 + }, + { + "epoch": 22.369009584664536, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 28006 + }, + { + "epoch": 22.369808306709267, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 28007 + }, + { + "epoch": 22.370607028753994, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 28008 + }, + { + "epoch": 22.371405750798722, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28009 + }, + { + "epoch": 22.37220447284345, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 28010 + }, + { + "epoch": 22.373003194888177, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 28011 + }, + { + "epoch": 22.37380191693291, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 28012 + }, + { + "epoch": 22.374600638977636, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 28013 + }, + { + "epoch": 22.375399361022364, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 28014 + }, + { + "epoch": 22.37619808306709, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 28015 + }, + { + "epoch": 22.376996805111823, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 28016 + }, + { + "epoch": 22.37779552715655, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 28017 + }, + { + "epoch": 22.378594249201278, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 28018 + }, + { + "epoch": 22.379392971246006, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 28019 + }, + { + "epoch": 22.380191693290733, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 28020 + }, + { + "epoch": 22.380990415335464, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 28021 + }, + { + "epoch": 22.381789137380192, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 28022 + }, + { + "epoch": 22.38258785942492, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 28023 + }, + { + "epoch": 22.383386581469647, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 28024 + }, + { + "epoch": 22.38418530351438, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 28025 + }, + { + "epoch": 22.384984025559106, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 28026 + }, + { + "epoch": 22.385782747603834, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 28027 + }, + { + "epoch": 22.38658146964856, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28028 + }, + { + "epoch": 22.38738019169329, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 28029 + }, + { + "epoch": 22.38817891373802, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28030 + }, + { + "epoch": 22.388977635782748, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 28031 + }, + { + "epoch": 22.389776357827476, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 28032 + }, + { + "epoch": 22.390575079872203, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 28033 + }, + { + "epoch": 22.391373801916934, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 28034 + }, + { + "epoch": 22.392172523961662, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 28035 + }, + { + "epoch": 22.39297124600639, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 28036 + }, + { + "epoch": 22.393769968051117, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 28037 + }, + { + "epoch": 22.394568690095845, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 28038 + }, + { + "epoch": 22.395367412140576, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 28039 + }, + { + "epoch": 22.396166134185304, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 28040 + }, + { + "epoch": 22.39696485623003, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 28041 + }, + { + "epoch": 22.39776357827476, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 28042 + }, + { + "epoch": 22.39856230031949, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 28043 + }, + { + "epoch": 22.399361022364218, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 28044 + }, + { + "epoch": 22.400159744408946, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 28045 + }, + { + "epoch": 22.400958466453673, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 28046 + }, + { + "epoch": 22.4017571884984, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 28047 + }, + { + "epoch": 22.402555910543132, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 28048 + }, + { + "epoch": 22.40335463258786, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 28049 + }, + { + "epoch": 22.404153354632587, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 28050 + }, + { + "epoch": 22.404952076677315, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 28051 + }, + { + "epoch": 22.405750798722046, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 28052 + }, + { + "epoch": 22.406549520766774, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 28053 + }, + { + "epoch": 22.4073482428115, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 28054 + }, + { + "epoch": 22.40814696485623, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 28055 + }, + { + "epoch": 22.408945686900957, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 28056 + }, + { + "epoch": 22.409744408945688, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 28057 + }, + { + "epoch": 22.410543130990416, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 28058 + }, + { + "epoch": 22.411341853035143, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 28059 + }, + { + "epoch": 22.41214057507987, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 28060 + }, + { + "epoch": 22.412939297124602, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 28061 + }, + { + "epoch": 22.41373801916933, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 28062 + }, + { + "epoch": 22.414536741214057, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 28063 + }, + { + "epoch": 22.415335463258785, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 28064 + }, + { + "epoch": 22.416134185303516, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 28065 + }, + { + "epoch": 22.416932907348244, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 28066 + }, + { + "epoch": 22.41773162939297, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 28067 + }, + { + "epoch": 22.4185303514377, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 28068 + }, + { + "epoch": 22.419329073482427, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0721, + "step": 28069 + }, + { + "epoch": 22.420127795527158, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 28070 + }, + { + "epoch": 22.420926517571885, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 28071 + }, + { + "epoch": 22.421725239616613, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 28072 + }, + { + "epoch": 22.42252396166134, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 28073 + }, + { + "epoch": 22.423322683706072, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 28074 + }, + { + "epoch": 22.4241214057508, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 28075 + }, + { + "epoch": 22.424920127795527, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 28076 + }, + { + "epoch": 22.425718849840255, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 28077 + }, + { + "epoch": 22.426517571884983, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 28078 + }, + { + "epoch": 22.427316293929714, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 28079 + }, + { + "epoch": 22.42811501597444, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 28080 + }, + { + "epoch": 22.42891373801917, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 28081 + }, + { + "epoch": 22.429712460063897, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 28082 + }, + { + "epoch": 22.430511182108628, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 28083 + }, + { + "epoch": 22.431309904153355, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 28084 + }, + { + "epoch": 22.432108626198083, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28085 + }, + { + "epoch": 22.43290734824281, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 28086 + }, + { + "epoch": 22.43370607028754, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 28087 + }, + { + "epoch": 22.43450479233227, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 28088 + }, + { + "epoch": 22.435303514376997, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 28089 + }, + { + "epoch": 22.436102236421725, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28090 + }, + { + "epoch": 22.436900958466452, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28091 + }, + { + "epoch": 22.437699680511184, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 28092 + }, + { + "epoch": 22.43849840255591, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 28093 + }, + { + "epoch": 22.43929712460064, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 28094 + }, + { + "epoch": 22.440095846645367, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 28095 + }, + { + "epoch": 22.440894568690094, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 28096 + }, + { + "epoch": 22.441693290734825, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 28097 + }, + { + "epoch": 22.442492012779553, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 28098 + }, + { + "epoch": 22.44329073482428, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 28099 + }, + { + "epoch": 22.44408945686901, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 28100 + }, + { + "epoch": 22.44488817891374, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 28101 + }, + { + "epoch": 22.445686900958467, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 28102 + }, + { + "epoch": 22.446485623003195, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 28103 + }, + { + "epoch": 22.447284345047922, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 28104 + }, + { + "epoch": 22.44808306709265, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 28105 + }, + { + "epoch": 22.44888178913738, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 28106 + }, + { + "epoch": 22.44968051118211, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 28107 + }, + { + "epoch": 22.450479233226837, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 28108 + }, + { + "epoch": 22.451277955271564, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 28109 + }, + { + "epoch": 22.452076677316295, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 28110 + }, + { + "epoch": 22.452875399361023, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 28111 + }, + { + "epoch": 22.45367412140575, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 28112 + }, + { + "epoch": 22.45447284345048, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 28113 + }, + { + "epoch": 22.455271565495206, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 28114 + }, + { + "epoch": 22.456070287539937, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 28115 + }, + { + "epoch": 22.456869009584665, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 28116 + }, + { + "epoch": 22.457667731629392, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 28117 + }, + { + "epoch": 22.45846645367412, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 28118 + }, + { + "epoch": 22.45926517571885, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 28119 + }, + { + "epoch": 22.46006389776358, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28120 + }, + { + "epoch": 22.460862619808307, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 28121 + }, + { + "epoch": 22.461661341853034, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 28122 + }, + { + "epoch": 22.462460063897762, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 28123 + }, + { + "epoch": 22.463258785942493, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 28124 + }, + { + "epoch": 22.46405750798722, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 28125 + }, + { + "epoch": 22.46485623003195, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 28126 + }, + { + "epoch": 22.465654952076676, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 28127 + }, + { + "epoch": 22.466453674121407, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28128 + }, + { + "epoch": 22.467252396166135, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 28129 + }, + { + "epoch": 22.468051118210862, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 28130 + }, + { + "epoch": 22.46884984025559, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28131 + }, + { + "epoch": 22.46964856230032, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 28132 + }, + { + "epoch": 22.47044728434505, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 28133 + }, + { + "epoch": 22.471246006389777, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 28134 + }, + { + "epoch": 22.472044728434504, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 28135 + }, + { + "epoch": 22.472843450479232, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 28136 + }, + { + "epoch": 22.473642172523963, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 28137 + }, + { + "epoch": 22.47444089456869, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 28138 + }, + { + "epoch": 22.47523961661342, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 28139 + }, + { + "epoch": 22.476038338658146, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 28140 + }, + { + "epoch": 22.476837060702877, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 28141 + }, + { + "epoch": 22.477635782747605, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 28142 + }, + { + "epoch": 22.478434504792332, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 28143 + }, + { + "epoch": 22.47923322683706, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 28144 + }, + { + "epoch": 22.480031948881788, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28145 + }, + { + "epoch": 22.48083067092652, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 28146 + }, + { + "epoch": 22.481629392971247, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 28147 + }, + { + "epoch": 22.482428115015974, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 28148 + }, + { + "epoch": 22.483226837060702, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 28149 + }, + { + "epoch": 22.484025559105433, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 28150 + }, + { + "epoch": 22.48482428115016, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 28151 + }, + { + "epoch": 22.48562300319489, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 28152 + }, + { + "epoch": 22.486421725239616, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 28153 + }, + { + "epoch": 22.487220447284344, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 28154 + }, + { + "epoch": 22.488019169329075, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 28155 + }, + { + "epoch": 22.488817891373802, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 28156 + }, + { + "epoch": 22.48961661341853, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 28157 + }, + { + "epoch": 22.490415335463258, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 28158 + }, + { + "epoch": 22.49121405750799, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 28159 + }, + { + "epoch": 22.492012779552716, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 28160 + }, + { + "epoch": 22.492811501597444, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28161 + }, + { + "epoch": 22.49361022364217, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 28162 + }, + { + "epoch": 22.4944089456869, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 28163 + }, + { + "epoch": 22.49520766773163, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 28164 + }, + { + "epoch": 22.49600638977636, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 28165 + }, + { + "epoch": 22.496805111821086, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 28166 + }, + { + "epoch": 22.497603833865814, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 28167 + }, + { + "epoch": 22.498402555910545, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28168 + }, + { + "epoch": 22.499201277955272, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 28169 + }, + { + "epoch": 22.5, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 28170 + }, + { + "epoch": 22.500798722044728, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 28171 + }, + { + "epoch": 22.501597444089455, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 28172 + }, + { + "epoch": 22.502396166134186, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 28173 + }, + { + "epoch": 22.503194888178914, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 28174 + }, + { + "epoch": 22.50399361022364, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 28175 + }, + { + "epoch": 22.50479233226837, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 28176 + }, + { + "epoch": 22.5055910543131, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 28177 + }, + { + "epoch": 22.50638977635783, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 28178 + }, + { + "epoch": 22.507188498402556, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 28179 + }, + { + "epoch": 22.507987220447284, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 28180 + }, + { + "epoch": 22.50878594249201, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 28181 + }, + { + "epoch": 22.509584664536742, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 28182 + }, + { + "epoch": 22.51038338658147, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 28183 + }, + { + "epoch": 22.511182108626198, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 28184 + }, + { + "epoch": 22.511980830670925, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 28185 + }, + { + "epoch": 22.512779552715656, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 28186 + }, + { + "epoch": 22.513578274760384, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 28187 + }, + { + "epoch": 22.51437699680511, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 28188 + }, + { + "epoch": 22.51517571884984, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 28189 + }, + { + "epoch": 22.515974440894567, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 28190 + }, + { + "epoch": 22.516773162939298, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 28191 + }, + { + "epoch": 22.517571884984026, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 28192 + }, + { + "epoch": 22.518370607028753, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 28193 + }, + { + "epoch": 22.51916932907348, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 28194 + }, + { + "epoch": 22.519968051118212, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 28195 + }, + { + "epoch": 22.52076677316294, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 28196 + }, + { + "epoch": 22.521565495207668, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 28197 + }, + { + "epoch": 22.522364217252395, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 28198 + }, + { + "epoch": 22.523162939297123, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 28199 + }, + { + "epoch": 22.523961661341854, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 28200 + }, + { + "epoch": 22.52476038338658, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 28201 + }, + { + "epoch": 22.52555910543131, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 28202 + }, + { + "epoch": 22.526357827476037, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28203 + }, + { + "epoch": 22.527156549520768, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 28204 + }, + { + "epoch": 22.527955271565496, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 28205 + }, + { + "epoch": 22.528753993610223, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 28206 + }, + { + "epoch": 22.52955271565495, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 28207 + }, + { + "epoch": 22.53035143769968, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 28208 + }, + { + "epoch": 22.53115015974441, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 28209 + }, + { + "epoch": 22.531948881789138, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 28210 + }, + { + "epoch": 22.532747603833865, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 28211 + }, + { + "epoch": 22.533546325878593, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 28212 + }, + { + "epoch": 22.534345047923324, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 28213 + }, + { + "epoch": 22.53514376996805, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 28214 + }, + { + "epoch": 22.53594249201278, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 28215 + }, + { + "epoch": 22.536741214057507, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 28216 + }, + { + "epoch": 22.537539936102235, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 28217 + }, + { + "epoch": 22.538338658146966, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 28218 + }, + { + "epoch": 22.539137380191693, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 28219 + }, + { + "epoch": 22.53993610223642, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 28220 + }, + { + "epoch": 22.54073482428115, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 28221 + }, + { + "epoch": 22.54153354632588, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 28222 + }, + { + "epoch": 22.542332268370608, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 28223 + }, + { + "epoch": 22.543130990415335, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 28224 + }, + { + "epoch": 22.543929712460063, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 28225 + }, + { + "epoch": 22.544728434504794, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 28226 + }, + { + "epoch": 22.54552715654952, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 28227 + }, + { + "epoch": 22.54632587859425, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 28228 + }, + { + "epoch": 22.547124600638977, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 28229 + }, + { + "epoch": 22.547923322683705, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 28230 + }, + { + "epoch": 22.548722044728436, + "grad_norm": 0.09765625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 28231 + }, + { + "epoch": 22.549520766773163, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 28232 + }, + { + "epoch": 22.55031948881789, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 28233 + }, + { + "epoch": 22.55111821086262, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28234 + }, + { + "epoch": 22.55191693290735, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 28235 + }, + { + "epoch": 22.552715654952078, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 28236 + }, + { + "epoch": 22.553514376996805, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 28237 + }, + { + "epoch": 22.554313099041533, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 28238 + }, + { + "epoch": 22.55511182108626, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28239 + }, + { + "epoch": 22.55591054313099, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 28240 + }, + { + "epoch": 22.55670926517572, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 28241 + }, + { + "epoch": 22.557507987220447, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28242 + }, + { + "epoch": 22.558306709265175, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 28243 + }, + { + "epoch": 22.559105431309906, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 28244 + }, + { + "epoch": 22.559904153354633, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 28245 + }, + { + "epoch": 22.56070287539936, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 28246 + }, + { + "epoch": 22.56150159744409, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 28247 + }, + { + "epoch": 22.562300319488816, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 28248 + }, + { + "epoch": 22.563099041533548, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 28249 + }, + { + "epoch": 22.563897763578275, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 28250 + }, + { + "epoch": 22.564696485623003, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 28251 + }, + { + "epoch": 22.56549520766773, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 28252 + }, + { + "epoch": 22.56629392971246, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 28253 + }, + { + "epoch": 22.56709265175719, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 28254 + }, + { + "epoch": 22.567891373801917, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 28255 + }, + { + "epoch": 22.568690095846645, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 28256 + }, + { + "epoch": 22.569488817891372, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 28257 + }, + { + "epoch": 22.570287539936103, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28258 + }, + { + "epoch": 22.57108626198083, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28259 + }, + { + "epoch": 22.57188498402556, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 28260 + }, + { + "epoch": 22.572683706070286, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 28261 + }, + { + "epoch": 22.573482428115017, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 28262 + }, + { + "epoch": 22.574281150159745, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 28263 + }, + { + "epoch": 22.575079872204473, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 28264 + }, + { + "epoch": 22.5758785942492, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 28265 + }, + { + "epoch": 22.576677316293928, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 28266 + }, + { + "epoch": 22.57747603833866, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 28267 + }, + { + "epoch": 22.578274760383387, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 28268 + }, + { + "epoch": 22.579073482428115, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 28269 + }, + { + "epoch": 22.579872204472842, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 28270 + }, + { + "epoch": 22.580670926517573, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 28271 + }, + { + "epoch": 22.5814696485623, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 28272 + }, + { + "epoch": 22.58226837060703, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 28273 + }, + { + "epoch": 22.583067092651756, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 28274 + }, + { + "epoch": 22.583865814696484, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 28275 + }, + { + "epoch": 22.584664536741215, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 28276 + }, + { + "epoch": 22.585463258785943, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 28277 + }, + { + "epoch": 22.58626198083067, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 28278 + }, + { + "epoch": 22.587060702875398, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 28279 + }, + { + "epoch": 22.58785942492013, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 28280 + }, + { + "epoch": 22.588658146964857, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 28281 + }, + { + "epoch": 22.589456869009584, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 28282 + }, + { + "epoch": 22.590255591054312, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 28283 + }, + { + "epoch": 22.591054313099043, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 28284 + }, + { + "epoch": 22.59185303514377, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 28285 + }, + { + "epoch": 22.5926517571885, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 28286 + }, + { + "epoch": 22.593450479233226, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 28287 + }, + { + "epoch": 22.594249201277954, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 28288 + }, + { + "epoch": 22.595047923322685, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 28289 + }, + { + "epoch": 22.595846645367413, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 28290 + }, + { + "epoch": 22.59664536741214, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 28291 + }, + { + "epoch": 22.597444089456868, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 28292 + }, + { + "epoch": 22.5982428115016, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 28293 + }, + { + "epoch": 22.599041533546327, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 28294 + }, + { + "epoch": 22.599840255591054, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 28295 + }, + { + "epoch": 22.600638977635782, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 28296 + }, + { + "epoch": 22.60143769968051, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 28297 + }, + { + "epoch": 22.60223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 28298 + }, + { + "epoch": 22.60303514376997, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 28299 + }, + { + "epoch": 22.603833865814696, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 28300 + }, + { + "epoch": 22.604632587859424, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 28301 + }, + { + "epoch": 22.605431309904155, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 28302 + }, + { + "epoch": 22.606230031948883, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 28303 + }, + { + "epoch": 22.60702875399361, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 28304 + }, + { + "epoch": 22.607827476038338, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 28305 + }, + { + "epoch": 22.608626198083066, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 28306 + }, + { + "epoch": 22.609424920127797, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 28307 + }, + { + "epoch": 22.610223642172524, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 28308 + }, + { + "epoch": 22.611022364217252, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 28309 + }, + { + "epoch": 22.61182108626198, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 28310 + }, + { + "epoch": 22.61261980830671, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28311 + }, + { + "epoch": 22.61341853035144, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28312 + }, + { + "epoch": 22.614217252396166, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 28313 + }, + { + "epoch": 22.615015974440894, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 28314 + }, + { + "epoch": 22.61581469648562, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28315 + }, + { + "epoch": 22.616613418530353, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 28316 + }, + { + "epoch": 22.61741214057508, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 28317 + }, + { + "epoch": 22.618210862619808, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 28318 + }, + { + "epoch": 22.619009584664536, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 28319 + }, + { + "epoch": 22.619808306709267, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 28320 + }, + { + "epoch": 22.620607028753994, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 28321 + }, + { + "epoch": 22.621405750798722, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 28322 + }, + { + "epoch": 22.62220447284345, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 28323 + }, + { + "epoch": 22.623003194888177, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 28324 + }, + { + "epoch": 22.62380191693291, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 28325 + }, + { + "epoch": 22.624600638977636, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 28326 + }, + { + "epoch": 22.625399361022364, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 28327 + }, + { + "epoch": 22.62619808306709, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 28328 + }, + { + "epoch": 22.626996805111823, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 28329 + }, + { + "epoch": 22.62779552715655, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 28330 + }, + { + "epoch": 22.628594249201278, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 28331 + }, + { + "epoch": 22.629392971246006, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 28332 + }, + { + "epoch": 22.630191693290733, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 28333 + }, + { + "epoch": 22.630990415335464, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 28334 + }, + { + "epoch": 22.631789137380192, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 28335 + }, + { + "epoch": 22.63258785942492, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 28336 + }, + { + "epoch": 22.633386581469647, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 28337 + }, + { + "epoch": 22.63418530351438, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28338 + }, + { + "epoch": 22.634984025559106, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 28339 + }, + { + "epoch": 22.635782747603834, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 28340 + }, + { + "epoch": 22.63658146964856, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 28341 + }, + { + "epoch": 22.63738019169329, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 28342 + }, + { + "epoch": 22.63817891373802, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 28343 + }, + { + "epoch": 22.638977635782748, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 28344 + }, + { + "epoch": 22.639776357827476, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 28345 + }, + { + "epoch": 22.640575079872203, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 28346 + }, + { + "epoch": 22.641373801916934, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 28347 + }, + { + "epoch": 22.642172523961662, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 28348 + }, + { + "epoch": 22.64297124600639, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 28349 + }, + { + "epoch": 22.643769968051117, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28350 + }, + { + "epoch": 22.644568690095845, + "grad_norm": 0.333984375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 28351 + }, + { + "epoch": 22.645367412140576, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 28352 + }, + { + "epoch": 22.646166134185304, + "grad_norm": 0.5390625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 28353 + }, + { + "epoch": 22.64696485623003, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 28354 + }, + { + "epoch": 22.64776357827476, + "grad_norm": 0.451171875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 28355 + }, + { + "epoch": 22.64856230031949, + "grad_norm": 0.314453125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28356 + }, + { + "epoch": 22.649361022364218, + "grad_norm": 0.5, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 28357 + }, + { + "epoch": 22.650159744408946, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 28358 + }, + { + "epoch": 22.650958466453673, + "grad_norm": 0.47265625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28359 + }, + { + "epoch": 22.6517571884984, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 28360 + }, + { + "epoch": 22.652555910543132, + "grad_norm": 0.353515625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 28361 + }, + { + "epoch": 22.65335463258786, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 28362 + }, + { + "epoch": 22.654153354632587, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28363 + }, + { + "epoch": 22.654952076677315, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28364 + }, + { + "epoch": 22.655750798722046, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 28365 + }, + { + "epoch": 22.656549520766774, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28366 + }, + { + "epoch": 22.6573482428115, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 28367 + }, + { + "epoch": 22.65814696485623, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 28368 + }, + { + "epoch": 22.658945686900957, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 28369 + }, + { + "epoch": 22.659744408945688, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 28370 + }, + { + "epoch": 22.660543130990416, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 28371 + }, + { + "epoch": 22.661341853035143, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 28372 + }, + { + "epoch": 22.66214057507987, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 28373 + }, + { + "epoch": 22.662939297124602, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 28374 + }, + { + "epoch": 22.66373801916933, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 28375 + }, + { + "epoch": 22.664536741214057, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 28376 + }, + { + "epoch": 22.665335463258785, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28377 + }, + { + "epoch": 22.666134185303516, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28378 + }, + { + "epoch": 22.666932907348244, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 28379 + }, + { + "epoch": 22.66773162939297, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 28380 + }, + { + "epoch": 22.6685303514377, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 28381 + }, + { + "epoch": 22.669329073482427, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 28382 + }, + { + "epoch": 22.670127795527158, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 28383 + }, + { + "epoch": 22.670926517571885, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 28384 + }, + { + "epoch": 22.671725239616613, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 28385 + }, + { + "epoch": 22.67252396166134, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 28386 + }, + { + "epoch": 22.673322683706072, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 28387 + }, + { + "epoch": 22.6741214057508, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 28388 + }, + { + "epoch": 22.674920127795527, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 28389 + }, + { + "epoch": 22.675718849840255, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 28390 + }, + { + "epoch": 22.676517571884983, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1073, + "step": 28391 + }, + { + "epoch": 22.677316293929714, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 28392 + }, + { + "epoch": 22.67811501597444, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 28393 + }, + { + "epoch": 22.67891373801917, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 28394 + }, + { + "epoch": 22.679712460063897, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 28395 + }, + { + "epoch": 22.680511182108628, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 28396 + }, + { + "epoch": 22.681309904153355, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 28397 + }, + { + "epoch": 22.682108626198083, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 28398 + }, + { + "epoch": 22.68290734824281, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 28399 + }, + { + "epoch": 22.68370607028754, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 28400 + }, + { + "epoch": 22.68450479233227, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 28401 + }, + { + "epoch": 22.685303514376997, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 28402 + }, + { + "epoch": 22.686102236421725, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 28403 + }, + { + "epoch": 22.686900958466452, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 28404 + }, + { + "epoch": 22.687699680511184, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 28405 + }, + { + "epoch": 22.68849840255591, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 28406 + }, + { + "epoch": 22.68929712460064, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 28407 + }, + { + "epoch": 22.690095846645367, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 28408 + }, + { + "epoch": 22.690894568690094, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 28409 + }, + { + "epoch": 22.691693290734825, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 28410 + }, + { + "epoch": 22.692492012779553, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 28411 + }, + { + "epoch": 22.69329073482428, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 28412 + }, + { + "epoch": 22.69408945686901, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 28413 + }, + { + "epoch": 22.69488817891374, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 28414 + }, + { + "epoch": 22.695686900958467, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 28415 + }, + { + "epoch": 22.696485623003195, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28416 + }, + { + "epoch": 22.697284345047922, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 28417 + }, + { + "epoch": 22.69808306709265, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 28418 + }, + { + "epoch": 22.69888178913738, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 28419 + }, + { + "epoch": 22.69968051118211, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28420 + }, + { + "epoch": 22.700479233226837, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 28421 + }, + { + "epoch": 22.701277955271564, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 28422 + }, + { + "epoch": 22.702076677316295, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 28423 + }, + { + "epoch": 22.702875399361023, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 28424 + }, + { + "epoch": 22.70367412140575, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 28425 + }, + { + "epoch": 22.70447284345048, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 28426 + }, + { + "epoch": 22.705271565495206, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 28427 + }, + { + "epoch": 22.706070287539937, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0752, + "step": 28428 + }, + { + "epoch": 22.706869009584665, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 28429 + }, + { + "epoch": 22.707667731629392, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 28430 + }, + { + "epoch": 22.70846645367412, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 28431 + }, + { + "epoch": 22.70926517571885, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 28432 + }, + { + "epoch": 22.71006389776358, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 28433 + }, + { + "epoch": 22.710862619808307, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 28434 + }, + { + "epoch": 22.711661341853034, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28435 + }, + { + "epoch": 22.712460063897765, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 28436 + }, + { + "epoch": 22.713258785942493, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 28437 + }, + { + "epoch": 22.71405750798722, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 28438 + }, + { + "epoch": 22.71485623003195, + "grad_norm": 0.33203125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 28439 + }, + { + "epoch": 22.715654952076676, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 28440 + }, + { + "epoch": 22.716453674121407, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 28441 + }, + { + "epoch": 22.717252396166135, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 28442 + }, + { + "epoch": 22.718051118210862, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 28443 + }, + { + "epoch": 22.71884984025559, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 28444 + }, + { + "epoch": 22.71964856230032, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 28445 + }, + { + "epoch": 22.72044728434505, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 28446 + }, + { + "epoch": 22.721246006389777, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 28447 + }, + { + "epoch": 22.722044728434504, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 28448 + }, + { + "epoch": 22.722843450479232, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 28449 + }, + { + "epoch": 22.723642172523963, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28450 + }, + { + "epoch": 22.72444089456869, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 28451 + }, + { + "epoch": 22.72523961661342, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 28452 + }, + { + "epoch": 22.726038338658146, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 28453 + }, + { + "epoch": 22.726837060702877, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 28454 + }, + { + "epoch": 22.727635782747605, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 28455 + }, + { + "epoch": 22.728434504792332, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 28456 + }, + { + "epoch": 22.72923322683706, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 28457 + }, + { + "epoch": 22.730031948881788, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 28458 + }, + { + "epoch": 22.73083067092652, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 28459 + }, + { + "epoch": 22.731629392971247, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 28460 + }, + { + "epoch": 22.732428115015974, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 28461 + }, + { + "epoch": 22.733226837060702, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 28462 + }, + { + "epoch": 22.734025559105433, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 28463 + }, + { + "epoch": 22.73482428115016, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 28464 + }, + { + "epoch": 22.73562300319489, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 28465 + }, + { + "epoch": 22.736421725239616, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 28466 + }, + { + "epoch": 22.737220447284344, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 28467 + }, + { + "epoch": 22.738019169329075, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 28468 + }, + { + "epoch": 22.738817891373802, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 28469 + }, + { + "epoch": 22.73961661341853, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 28470 + }, + { + "epoch": 22.740415335463258, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 28471 + }, + { + "epoch": 22.74121405750799, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 28472 + }, + { + "epoch": 22.742012779552716, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 28473 + }, + { + "epoch": 22.742811501597444, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 28474 + }, + { + "epoch": 22.74361022364217, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 28475 + }, + { + "epoch": 22.7444089456869, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 28476 + }, + { + "epoch": 22.74520766773163, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 28477 + }, + { + "epoch": 22.74600638977636, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 28478 + }, + { + "epoch": 22.746805111821086, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 28479 + }, + { + "epoch": 22.747603833865814, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 28480 + }, + { + "epoch": 22.748402555910545, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 28481 + }, + { + "epoch": 22.749201277955272, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 28482 + }, + { + "epoch": 22.75, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 28483 + }, + { + "epoch": 22.750798722044728, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 28484 + }, + { + "epoch": 22.751597444089455, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 28485 + }, + { + "epoch": 22.752396166134186, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 28486 + }, + { + "epoch": 22.753194888178914, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 28487 + }, + { + "epoch": 22.75399361022364, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28488 + }, + { + "epoch": 22.75479233226837, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 28489 + }, + { + "epoch": 22.7555910543131, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28490 + }, + { + "epoch": 22.75638977635783, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 28491 + }, + { + "epoch": 22.757188498402556, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 28492 + }, + { + "epoch": 22.757987220447284, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 28493 + }, + { + "epoch": 22.75878594249201, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 28494 + }, + { + "epoch": 22.759584664536742, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 28495 + }, + { + "epoch": 22.76038338658147, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 28496 + }, + { + "epoch": 22.761182108626198, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 28497 + }, + { + "epoch": 22.761980830670925, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 28498 + }, + { + "epoch": 22.762779552715656, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 28499 + }, + { + "epoch": 22.763578274760384, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 28500 + }, + { + "epoch": 22.76437699680511, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 28501 + }, + { + "epoch": 22.76517571884984, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 28502 + }, + { + "epoch": 22.765974440894567, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 28503 + }, + { + "epoch": 22.766773162939298, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 28504 + }, + { + "epoch": 22.767571884984026, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 28505 + }, + { + "epoch": 22.768370607028753, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 28506 + }, + { + "epoch": 22.76916932907348, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 28507 + }, + { + "epoch": 22.769968051118212, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 28508 + }, + { + "epoch": 22.77076677316294, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 28509 + }, + { + "epoch": 22.771565495207668, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 28510 + }, + { + "epoch": 22.772364217252395, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 28511 + }, + { + "epoch": 22.773162939297123, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 28512 + }, + { + "epoch": 22.773961661341854, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 28513 + }, + { + "epoch": 22.77476038338658, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 28514 + }, + { + "epoch": 22.77555910543131, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1055, + "step": 28515 + }, + { + "epoch": 22.776357827476037, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 28516 + }, + { + "epoch": 22.777156549520768, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 28517 + }, + { + "epoch": 22.777955271565496, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 28518 + }, + { + "epoch": 22.778753993610223, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 28519 + }, + { + "epoch": 22.77955271565495, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 28520 + }, + { + "epoch": 22.78035143769968, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 28521 + }, + { + "epoch": 22.78115015974441, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 28522 + }, + { + "epoch": 22.781948881789138, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 28523 + }, + { + "epoch": 22.782747603833865, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 28524 + }, + { + "epoch": 22.783546325878593, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 28525 + }, + { + "epoch": 22.784345047923324, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 28526 + }, + { + "epoch": 22.78514376996805, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 28527 + }, + { + "epoch": 22.78594249201278, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 28528 + }, + { + "epoch": 22.786741214057507, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 28529 + }, + { + "epoch": 22.787539936102235, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 28530 + }, + { + "epoch": 22.788338658146966, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 28531 + }, + { + "epoch": 22.789137380191693, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 28532 + }, + { + "epoch": 22.78993610223642, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 28533 + }, + { + "epoch": 22.79073482428115, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 28534 + }, + { + "epoch": 22.79153354632588, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 28535 + }, + { + "epoch": 22.792332268370608, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 28536 + }, + { + "epoch": 22.793130990415335, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 28537 + }, + { + "epoch": 22.793929712460063, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 28538 + }, + { + "epoch": 22.794728434504794, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 28539 + }, + { + "epoch": 22.79552715654952, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 28540 + }, + { + "epoch": 22.79632587859425, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 28541 + }, + { + "epoch": 22.797124600638977, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 28542 + }, + { + "epoch": 22.797923322683705, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 28543 + }, + { + "epoch": 22.798722044728436, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 28544 + }, + { + "epoch": 22.799520766773163, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 28545 + }, + { + "epoch": 22.80031948881789, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 28546 + }, + { + "epoch": 22.80111821086262, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 28547 + }, + { + "epoch": 22.80191693290735, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 28548 + }, + { + "epoch": 22.802715654952078, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 28549 + }, + { + "epoch": 22.803514376996805, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 28550 + }, + { + "epoch": 22.804313099041533, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 28551 + }, + { + "epoch": 22.80511182108626, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 28552 + }, + { + "epoch": 22.80591054313099, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 28553 + }, + { + "epoch": 22.80670926517572, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 28554 + }, + { + "epoch": 22.807507987220447, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 28555 + }, + { + "epoch": 22.808306709265175, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 28556 + }, + { + "epoch": 22.809105431309906, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 28557 + }, + { + "epoch": 22.809904153354633, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 28558 + }, + { + "epoch": 22.81070287539936, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 28559 + }, + { + "epoch": 22.81150159744409, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 28560 + }, + { + "epoch": 22.812300319488816, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 28561 + }, + { + "epoch": 22.813099041533548, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 28562 + }, + { + "epoch": 22.813897763578275, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 28563 + }, + { + "epoch": 22.814696485623003, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 28564 + }, + { + "epoch": 22.81549520766773, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 28565 + }, + { + "epoch": 22.81629392971246, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 28566 + }, + { + "epoch": 22.81709265175719, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 28567 + }, + { + "epoch": 22.817891373801917, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 28568 + }, + { + "epoch": 22.818690095846645, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 28569 + }, + { + "epoch": 22.819488817891372, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 28570 + }, + { + "epoch": 22.820287539936103, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 28571 + }, + { + "epoch": 22.82108626198083, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 28572 + }, + { + "epoch": 22.82188498402556, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 28573 + }, + { + "epoch": 22.822683706070286, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 28574 + }, + { + "epoch": 22.823482428115017, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 28575 + }, + { + "epoch": 22.824281150159745, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 28576 + }, + { + "epoch": 22.825079872204473, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 28577 + }, + { + "epoch": 22.8258785942492, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 28578 + }, + { + "epoch": 22.826677316293928, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 28579 + }, + { + "epoch": 22.82747603833866, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 28580 + }, + { + "epoch": 22.828274760383387, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 28581 + }, + { + "epoch": 22.829073482428115, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 28582 + }, + { + "epoch": 22.829872204472842, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 28583 + }, + { + "epoch": 22.830670926517573, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 28584 + }, + { + "epoch": 22.8314696485623, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28585 + }, + { + "epoch": 22.83226837060703, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 28586 + }, + { + "epoch": 22.833067092651756, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 28587 + }, + { + "epoch": 22.833865814696484, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28588 + }, + { + "epoch": 22.834664536741215, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 28589 + }, + { + "epoch": 22.835463258785943, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 28590 + }, + { + "epoch": 22.83626198083067, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 28591 + }, + { + "epoch": 22.837060702875398, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 28592 + }, + { + "epoch": 22.83785942492013, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 28593 + }, + { + "epoch": 22.838658146964857, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 28594 + }, + { + "epoch": 22.839456869009584, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28595 + }, + { + "epoch": 22.840255591054312, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 28596 + }, + { + "epoch": 22.841054313099043, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 28597 + }, + { + "epoch": 22.84185303514377, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 28598 + }, + { + "epoch": 22.8426517571885, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 28599 + }, + { + "epoch": 22.843450479233226, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 28600 + }, + { + "epoch": 22.844249201277954, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 28601 + }, + { + "epoch": 22.845047923322685, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 28602 + }, + { + "epoch": 22.845846645367413, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 28603 + }, + { + "epoch": 22.84664536741214, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 28604 + }, + { + "epoch": 22.847444089456868, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 28605 + }, + { + "epoch": 22.8482428115016, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 28606 + }, + { + "epoch": 22.849041533546327, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 28607 + }, + { + "epoch": 22.849840255591054, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 28608 + }, + { + "epoch": 22.850638977635782, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 28609 + }, + { + "epoch": 22.85143769968051, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 28610 + }, + { + "epoch": 22.85223642172524, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 28611 + }, + { + "epoch": 22.85303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 28612 + }, + { + "epoch": 22.853833865814696, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 28613 + }, + { + "epoch": 22.854632587859424, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 28614 + }, + { + "epoch": 22.855431309904155, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 28615 + }, + { + "epoch": 22.856230031948883, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1051, + "step": 28616 + }, + { + "epoch": 22.85702875399361, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 28617 + }, + { + "epoch": 22.857827476038338, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 28618 + }, + { + "epoch": 22.858626198083066, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 28619 + }, + { + "epoch": 22.859424920127797, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 28620 + }, + { + "epoch": 22.860223642172524, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 28621 + }, + { + "epoch": 22.861022364217252, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 28622 + }, + { + "epoch": 22.86182108626198, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 28623 + }, + { + "epoch": 22.86261980830671, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 28624 + }, + { + "epoch": 22.86341853035144, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 28625 + }, + { + "epoch": 22.864217252396166, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 28626 + }, + { + "epoch": 22.865015974440894, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 28627 + }, + { + "epoch": 22.86581469648562, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 28628 + }, + { + "epoch": 22.866613418530353, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 28629 + }, + { + "epoch": 22.86741214057508, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 28630 + }, + { + "epoch": 22.868210862619808, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 28631 + }, + { + "epoch": 22.869009584664536, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 28632 + }, + { + "epoch": 22.869808306709267, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 28633 + }, + { + "epoch": 22.870607028753994, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 28634 + }, + { + "epoch": 22.871405750798722, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 28635 + }, + { + "epoch": 22.87220447284345, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 28636 + }, + { + "epoch": 22.873003194888177, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 28637 + }, + { + "epoch": 22.87380191693291, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 28638 + }, + { + "epoch": 22.874600638977636, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 28639 + }, + { + "epoch": 22.875399361022364, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 28640 + }, + { + "epoch": 22.87619808306709, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 28641 + }, + { + "epoch": 22.876996805111823, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 28642 + }, + { + "epoch": 22.87779552715655, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 28643 + }, + { + "epoch": 22.878594249201278, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 28644 + }, + { + "epoch": 22.879392971246006, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 28645 + }, + { + "epoch": 22.880191693290733, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 28646 + }, + { + "epoch": 22.880990415335464, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 28647 + }, + { + "epoch": 22.881789137380192, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 28648 + }, + { + "epoch": 22.88258785942492, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 28649 + }, + { + "epoch": 22.883386581469647, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 28650 + }, + { + "epoch": 22.88418530351438, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 28651 + }, + { + "epoch": 22.884984025559106, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 28652 + }, + { + "epoch": 22.885782747603834, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 28653 + }, + { + "epoch": 22.88658146964856, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 28654 + }, + { + "epoch": 22.88738019169329, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 28655 + }, + { + "epoch": 22.88817891373802, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 28656 + }, + { + "epoch": 22.888977635782748, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 28657 + }, + { + "epoch": 22.889776357827476, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 28658 + }, + { + "epoch": 22.890575079872203, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 28659 + }, + { + "epoch": 22.891373801916934, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 28660 + }, + { + "epoch": 22.892172523961662, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 28661 + }, + { + "epoch": 22.89297124600639, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 28662 + }, + { + "epoch": 22.893769968051117, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 28663 + }, + { + "epoch": 22.894568690095845, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 28664 + }, + { + "epoch": 22.895367412140576, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 28665 + }, + { + "epoch": 22.896166134185304, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 28666 + }, + { + "epoch": 22.89696485623003, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 28667 + }, + { + "epoch": 22.89776357827476, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 28668 + }, + { + "epoch": 22.89856230031949, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 28669 + }, + { + "epoch": 22.899361022364218, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28670 + }, + { + "epoch": 22.900159744408946, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 28671 + }, + { + "epoch": 22.900958466453673, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 28672 + }, + { + "epoch": 22.9017571884984, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 28673 + }, + { + "epoch": 22.902555910543132, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 28674 + }, + { + "epoch": 22.90335463258786, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28675 + }, + { + "epoch": 22.904153354632587, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 28676 + }, + { + "epoch": 22.904952076677315, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 28677 + }, + { + "epoch": 22.905750798722046, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 28678 + }, + { + "epoch": 22.906549520766774, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 28679 + }, + { + "epoch": 22.9073482428115, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 28680 + }, + { + "epoch": 22.90814696485623, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 28681 + }, + { + "epoch": 22.908945686900957, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 28682 + }, + { + "epoch": 22.909744408945688, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 28683 + }, + { + "epoch": 22.910543130990416, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 28684 + }, + { + "epoch": 22.911341853035143, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 28685 + }, + { + "epoch": 22.91214057507987, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 28686 + }, + { + "epoch": 22.912939297124602, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 28687 + }, + { + "epoch": 22.91373801916933, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 28688 + }, + { + "epoch": 22.914536741214057, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 28689 + }, + { + "epoch": 22.915335463258785, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 28690 + }, + { + "epoch": 22.916134185303516, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 28691 + }, + { + "epoch": 22.916932907348244, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 28692 + }, + { + "epoch": 22.91773162939297, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 28693 + }, + { + "epoch": 22.9185303514377, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 28694 + }, + { + "epoch": 22.919329073482427, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 28695 + }, + { + "epoch": 22.920127795527158, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 28696 + }, + { + "epoch": 22.920926517571885, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 28697 + }, + { + "epoch": 22.921725239616613, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 28698 + }, + { + "epoch": 22.92252396166134, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 28699 + }, + { + "epoch": 22.923322683706072, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 28700 + }, + { + "epoch": 22.9241214057508, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 28701 + }, + { + "epoch": 22.924920127795527, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 28702 + }, + { + "epoch": 22.925718849840255, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 28703 + }, + { + "epoch": 22.926517571884983, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 28704 + }, + { + "epoch": 22.927316293929714, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 28705 + }, + { + "epoch": 22.92811501597444, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 28706 + }, + { + "epoch": 22.92891373801917, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28707 + }, + { + "epoch": 22.929712460063897, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 28708 + }, + { + "epoch": 22.930511182108628, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 28709 + }, + { + "epoch": 22.931309904153355, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 28710 + }, + { + "epoch": 22.932108626198083, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 28711 + }, + { + "epoch": 22.93290734824281, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 28712 + }, + { + "epoch": 22.93370607028754, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 28713 + }, + { + "epoch": 22.93450479233227, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 28714 + }, + { + "epoch": 22.935303514376997, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 28715 + }, + { + "epoch": 22.936102236421725, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28716 + }, + { + "epoch": 22.936900958466452, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 28717 + }, + { + "epoch": 22.937699680511184, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 28718 + }, + { + "epoch": 22.93849840255591, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 28719 + }, + { + "epoch": 22.93929712460064, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 28720 + }, + { + "epoch": 22.940095846645367, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 28721 + }, + { + "epoch": 22.940894568690094, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 28722 + }, + { + "epoch": 22.941693290734825, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 28723 + }, + { + "epoch": 22.942492012779553, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 28724 + }, + { + "epoch": 22.94329073482428, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 28725 + }, + { + "epoch": 22.94408945686901, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 28726 + }, + { + "epoch": 22.94488817891374, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.1061, + "step": 28727 + }, + { + "epoch": 22.945686900958467, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 28728 + }, + { + "epoch": 22.946485623003195, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 28729 + }, + { + "epoch": 22.947284345047922, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28730 + }, + { + "epoch": 22.94808306709265, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 28731 + }, + { + "epoch": 22.94888178913738, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 28732 + }, + { + "epoch": 22.94968051118211, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28733 + }, + { + "epoch": 22.950479233226837, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 28734 + }, + { + "epoch": 22.951277955271564, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 28735 + }, + { + "epoch": 22.952076677316295, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 28736 + }, + { + "epoch": 22.952875399361023, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 28737 + }, + { + "epoch": 22.95367412140575, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 28738 + }, + { + "epoch": 22.95447284345048, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 28739 + }, + { + "epoch": 22.955271565495206, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 28740 + }, + { + "epoch": 22.956070287539937, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 28741 + }, + { + "epoch": 22.956869009584665, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 28742 + }, + { + "epoch": 22.957667731629392, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 28743 + }, + { + "epoch": 22.95846645367412, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 28744 + }, + { + "epoch": 22.95926517571885, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28745 + }, + { + "epoch": 22.96006389776358, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 28746 + }, + { + "epoch": 22.960862619808307, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 28747 + }, + { + "epoch": 22.961661341853034, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 28748 + }, + { + "epoch": 22.962460063897765, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 28749 + }, + { + "epoch": 22.963258785942493, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 28750 + }, + { + "epoch": 22.96405750798722, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28751 + }, + { + "epoch": 22.96485623003195, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 28752 + }, + { + "epoch": 22.965654952076676, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 28753 + }, + { + "epoch": 22.966453674121407, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 28754 + }, + { + "epoch": 22.967252396166135, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 28755 + }, + { + "epoch": 22.968051118210862, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 28756 + }, + { + "epoch": 22.96884984025559, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28757 + }, + { + "epoch": 22.96964856230032, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 28758 + }, + { + "epoch": 22.97044728434505, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 28759 + }, + { + "epoch": 22.971246006389777, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 28760 + }, + { + "epoch": 22.972044728434504, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 28761 + }, + { + "epoch": 22.972843450479232, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 28762 + }, + { + "epoch": 22.973642172523963, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 28763 + }, + { + "epoch": 22.97444089456869, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 28764 + }, + { + "epoch": 22.97523961661342, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 28765 + }, + { + "epoch": 22.976038338658146, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 28766 + }, + { + "epoch": 22.976837060702877, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 28767 + }, + { + "epoch": 22.977635782747605, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 28768 + }, + { + "epoch": 22.978434504792332, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28769 + }, + { + "epoch": 22.97923322683706, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28770 + }, + { + "epoch": 22.980031948881788, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 28771 + }, + { + "epoch": 22.98083067092652, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 28772 + }, + { + "epoch": 22.981629392971247, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 28773 + }, + { + "epoch": 22.982428115015974, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28774 + }, + { + "epoch": 22.983226837060702, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 28775 + }, + { + "epoch": 22.984025559105433, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 28776 + }, + { + "epoch": 22.98482428115016, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 28777 + }, + { + "epoch": 22.98562300319489, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 28778 + }, + { + "epoch": 22.986421725239616, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 28779 + }, + { + "epoch": 22.987220447284344, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 28780 + }, + { + "epoch": 22.988019169329075, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 28781 + }, + { + "epoch": 22.988817891373802, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 28782 + }, + { + "epoch": 22.98961661341853, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 28783 + }, + { + "epoch": 22.990415335463258, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 28784 + }, + { + "epoch": 22.99121405750799, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 28785 + }, + { + "epoch": 22.992012779552716, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 28786 + }, + { + "epoch": 22.992811501597444, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 28787 + }, + { + "epoch": 22.99361022364217, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 28788 + }, + { + "epoch": 22.9944089456869, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 28789 + }, + { + "epoch": 22.99520766773163, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 28790 + }, + { + "epoch": 22.99600638977636, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 28791 + }, + { + "epoch": 22.996805111821086, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 28792 + }, + { + "epoch": 22.997603833865814, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 28793 + }, + { + "epoch": 22.998402555910545, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 28794 + }, + { + "epoch": 22.999201277955272, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 28795 + }, + { + "epoch": 23.0, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 28796 + }, + { + "epoch": 23.000798722044728, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 28797 + }, + { + "epoch": 23.001597444089455, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28798 + }, + { + "epoch": 23.002396166134186, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28799 + }, + { + "epoch": 23.003194888178914, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 28800 + }, + { + "epoch": 23.00399361022364, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 28801 + }, + { + "epoch": 23.00479233226837, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 28802 + }, + { + "epoch": 23.0055910543131, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 28803 + }, + { + "epoch": 23.00638977635783, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 28804 + }, + { + "epoch": 23.007188498402556, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 28805 + }, + { + "epoch": 23.007987220447284, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 28806 + }, + { + "epoch": 23.00878594249201, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 28807 + }, + { + "epoch": 23.009584664536742, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 28808 + }, + { + "epoch": 23.01038338658147, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 28809 + }, + { + "epoch": 23.011182108626198, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 28810 + }, + { + "epoch": 23.011980830670925, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 28811 + }, + { + "epoch": 23.012779552715656, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 28812 + }, + { + "epoch": 23.013578274760384, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 28813 + }, + { + "epoch": 23.01437699680511, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 28814 + }, + { + "epoch": 23.01517571884984, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 28815 + }, + { + "epoch": 23.015974440894567, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 28816 + }, + { + "epoch": 23.016773162939298, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28817 + }, + { + "epoch": 23.017571884984026, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 28818 + }, + { + "epoch": 23.018370607028753, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 28819 + }, + { + "epoch": 23.01916932907348, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 28820 + }, + { + "epoch": 23.019968051118212, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 28821 + }, + { + "epoch": 23.02076677316294, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 28822 + }, + { + "epoch": 23.021565495207668, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 28823 + }, + { + "epoch": 23.022364217252395, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 28824 + }, + { + "epoch": 23.023162939297123, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 28825 + }, + { + "epoch": 23.023961661341854, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 28826 + }, + { + "epoch": 23.02476038338658, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 28827 + }, + { + "epoch": 23.02555910543131, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 28828 + }, + { + "epoch": 23.026357827476037, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 28829 + }, + { + "epoch": 23.027156549520768, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 28830 + }, + { + "epoch": 23.027955271565496, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 28831 + }, + { + "epoch": 23.028753993610223, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 28832 + }, + { + "epoch": 23.02955271565495, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 28833 + }, + { + "epoch": 23.03035143769968, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 28834 + }, + { + "epoch": 23.03115015974441, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 28835 + }, + { + "epoch": 23.031948881789138, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28836 + }, + { + "epoch": 23.032747603833865, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 28837 + }, + { + "epoch": 23.033546325878593, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 28838 + }, + { + "epoch": 23.034345047923324, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 28839 + }, + { + "epoch": 23.03514376996805, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.1017, + "step": 28840 + }, + { + "epoch": 23.03594249201278, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 28841 + }, + { + "epoch": 23.036741214057507, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 28842 + }, + { + "epoch": 23.037539936102238, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28843 + }, + { + "epoch": 23.038338658146966, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 28844 + }, + { + "epoch": 23.039137380191693, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 28845 + }, + { + "epoch": 23.03993610223642, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 28846 + }, + { + "epoch": 23.04073482428115, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 28847 + }, + { + "epoch": 23.04153354632588, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 28848 + }, + { + "epoch": 23.042332268370608, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 28849 + }, + { + "epoch": 23.043130990415335, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 28850 + }, + { + "epoch": 23.043929712460063, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28851 + }, + { + "epoch": 23.044728434504794, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 28852 + }, + { + "epoch": 23.04552715654952, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 28853 + }, + { + "epoch": 23.04632587859425, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 28854 + }, + { + "epoch": 23.047124600638977, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 28855 + }, + { + "epoch": 23.047923322683705, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 28856 + }, + { + "epoch": 23.048722044728436, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 28857 + }, + { + "epoch": 23.049520766773163, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 28858 + }, + { + "epoch": 23.05031948881789, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 28859 + }, + { + "epoch": 23.05111821086262, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 28860 + }, + { + "epoch": 23.05191693290735, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 28861 + }, + { + "epoch": 23.052715654952078, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 28862 + }, + { + "epoch": 23.053514376996805, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 28863 + }, + { + "epoch": 23.054313099041533, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 28864 + }, + { + "epoch": 23.05511182108626, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 28865 + }, + { + "epoch": 23.05591054313099, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 28866 + }, + { + "epoch": 23.05670926517572, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 28867 + }, + { + "epoch": 23.057507987220447, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 28868 + }, + { + "epoch": 23.058306709265175, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 28869 + }, + { + "epoch": 23.059105431309906, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 28870 + }, + { + "epoch": 23.059904153354633, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 28871 + }, + { + "epoch": 23.06070287539936, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 28872 + }, + { + "epoch": 23.06150159744409, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 28873 + }, + { + "epoch": 23.062300319488816, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 28874 + }, + { + "epoch": 23.063099041533548, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 28875 + }, + { + "epoch": 23.063897763578275, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 28876 + }, + { + "epoch": 23.064696485623003, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 28877 + }, + { + "epoch": 23.06549520766773, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 28878 + }, + { + "epoch": 23.06629392971246, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 28879 + }, + { + "epoch": 23.06709265175719, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28880 + }, + { + "epoch": 23.067891373801917, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28881 + }, + { + "epoch": 23.068690095846645, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 28882 + }, + { + "epoch": 23.069488817891372, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 28883 + }, + { + "epoch": 23.070287539936103, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 28884 + }, + { + "epoch": 23.07108626198083, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 28885 + }, + { + "epoch": 23.07188498402556, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 28886 + }, + { + "epoch": 23.072683706070286, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 28887 + }, + { + "epoch": 23.073482428115017, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 28888 + }, + { + "epoch": 23.074281150159745, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28889 + }, + { + "epoch": 23.075079872204473, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 28890 + }, + { + "epoch": 23.0758785942492, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 28891 + }, + { + "epoch": 23.076677316293928, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 28892 + }, + { + "epoch": 23.07747603833866, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 28893 + }, + { + "epoch": 23.078274760383387, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 28894 + }, + { + "epoch": 23.079073482428115, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 28895 + }, + { + "epoch": 23.079872204472842, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 28896 + }, + { + "epoch": 23.080670926517573, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 28897 + }, + { + "epoch": 23.0814696485623, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28898 + }, + { + "epoch": 23.08226837060703, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 28899 + }, + { + "epoch": 23.083067092651756, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 28900 + }, + { + "epoch": 23.083865814696484, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 28901 + }, + { + "epoch": 23.084664536741215, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 28902 + }, + { + "epoch": 23.085463258785943, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 28903 + }, + { + "epoch": 23.08626198083067, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 28904 + }, + { + "epoch": 23.087060702875398, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 28905 + }, + { + "epoch": 23.08785942492013, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 28906 + }, + { + "epoch": 23.088658146964857, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 28907 + }, + { + "epoch": 23.089456869009584, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 28908 + }, + { + "epoch": 23.090255591054312, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 28909 + }, + { + "epoch": 23.09105431309904, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 28910 + }, + { + "epoch": 23.09185303514377, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 28911 + }, + { + "epoch": 23.0926517571885, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 28912 + }, + { + "epoch": 23.093450479233226, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 28913 + }, + { + "epoch": 23.094249201277954, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 28914 + }, + { + "epoch": 23.095047923322685, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 28915 + }, + { + "epoch": 23.095846645367413, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 28916 + }, + { + "epoch": 23.09664536741214, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 28917 + }, + { + "epoch": 23.097444089456868, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 28918 + }, + { + "epoch": 23.0982428115016, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 28919 + }, + { + "epoch": 23.099041533546327, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 28920 + }, + { + "epoch": 23.099840255591054, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 28921 + }, + { + "epoch": 23.100638977635782, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 28922 + }, + { + "epoch": 23.10143769968051, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 28923 + }, + { + "epoch": 23.10223642172524, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 28924 + }, + { + "epoch": 23.10303514376997, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 28925 + }, + { + "epoch": 23.103833865814696, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 28926 + }, + { + "epoch": 23.104632587859424, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 28927 + }, + { + "epoch": 23.105431309904155, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 28928 + }, + { + "epoch": 23.106230031948883, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 28929 + }, + { + "epoch": 23.10702875399361, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 28930 + }, + { + "epoch": 23.107827476038338, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 28931 + }, + { + "epoch": 23.108626198083066, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 28932 + }, + { + "epoch": 23.109424920127797, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 28933 + }, + { + "epoch": 23.110223642172524, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 28934 + }, + { + "epoch": 23.111022364217252, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 28935 + }, + { + "epoch": 23.11182108626198, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 28936 + }, + { + "epoch": 23.11261980830671, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 28937 + }, + { + "epoch": 23.11341853035144, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 28938 + }, + { + "epoch": 23.114217252396166, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 28939 + }, + { + "epoch": 23.115015974440894, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 28940 + }, + { + "epoch": 23.11581469648562, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 28941 + }, + { + "epoch": 23.116613418530353, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 28942 + }, + { + "epoch": 23.11741214057508, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 28943 + }, + { + "epoch": 23.118210862619808, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 28944 + }, + { + "epoch": 23.119009584664536, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 28945 + }, + { + "epoch": 23.119808306709267, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 28946 + }, + { + "epoch": 23.120607028753994, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 28947 + }, + { + "epoch": 23.121405750798722, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 28948 + }, + { + "epoch": 23.12220447284345, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 28949 + }, + { + "epoch": 23.123003194888177, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 28950 + }, + { + "epoch": 23.12380191693291, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 28951 + }, + { + "epoch": 23.124600638977636, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28952 + }, + { + "epoch": 23.125399361022364, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 28953 + }, + { + "epoch": 23.12619808306709, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 28954 + }, + { + "epoch": 23.126996805111823, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 28955 + }, + { + "epoch": 23.12779552715655, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 28956 + }, + { + "epoch": 23.128594249201278, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 28957 + }, + { + "epoch": 23.129392971246006, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 28958 + }, + { + "epoch": 23.130191693290733, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 28959 + }, + { + "epoch": 23.130990415335464, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 28960 + }, + { + "epoch": 23.131789137380192, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 28961 + }, + { + "epoch": 23.13258785942492, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 28962 + }, + { + "epoch": 23.133386581469647, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 28963 + }, + { + "epoch": 23.13418530351438, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 28964 + }, + { + "epoch": 23.134984025559106, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 28965 + }, + { + "epoch": 23.135782747603834, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 28966 + }, + { + "epoch": 23.13658146964856, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 28967 + }, + { + "epoch": 23.13738019169329, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 28968 + }, + { + "epoch": 23.13817891373802, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 28969 + }, + { + "epoch": 23.138977635782748, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 28970 + }, + { + "epoch": 23.139776357827476, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 28971 + }, + { + "epoch": 23.140575079872203, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 28972 + }, + { + "epoch": 23.141373801916934, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 28973 + }, + { + "epoch": 23.142172523961662, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 28974 + }, + { + "epoch": 23.14297124600639, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 28975 + }, + { + "epoch": 23.143769968051117, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 28976 + }, + { + "epoch": 23.144568690095845, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 28977 + }, + { + "epoch": 23.145367412140576, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 28978 + }, + { + "epoch": 23.146166134185304, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 28979 + }, + { + "epoch": 23.14696485623003, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 28980 + }, + { + "epoch": 23.14776357827476, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 28981 + }, + { + "epoch": 23.14856230031949, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 28982 + }, + { + "epoch": 23.149361022364218, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 28983 + }, + { + "epoch": 23.150159744408946, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 28984 + }, + { + "epoch": 23.150958466453673, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 28985 + }, + { + "epoch": 23.1517571884984, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 28986 + }, + { + "epoch": 23.152555910543132, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 28987 + }, + { + "epoch": 23.15335463258786, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 28988 + }, + { + "epoch": 23.154153354632587, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 28989 + }, + { + "epoch": 23.154952076677315, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 28990 + }, + { + "epoch": 23.155750798722046, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 28991 + }, + { + "epoch": 23.156549520766774, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 28992 + }, + { + "epoch": 23.1573482428115, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 28993 + }, + { + "epoch": 23.15814696485623, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 28994 + }, + { + "epoch": 23.15894568690096, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 28995 + }, + { + "epoch": 23.159744408945688, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 28996 + }, + { + "epoch": 23.160543130990416, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 28997 + }, + { + "epoch": 23.161341853035143, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 28998 + }, + { + "epoch": 23.16214057507987, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 28999 + }, + { + "epoch": 23.162939297124602, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 29000 + }, + { + "epoch": 23.16373801916933, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 29001 + }, + { + "epoch": 23.164536741214057, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29002 + }, + { + "epoch": 23.165335463258785, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 29003 + }, + { + "epoch": 23.166134185303516, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 29004 + }, + { + "epoch": 23.166932907348244, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 29005 + }, + { + "epoch": 23.16773162939297, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 29006 + }, + { + "epoch": 23.1685303514377, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 29007 + }, + { + "epoch": 23.169329073482427, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 29008 + }, + { + "epoch": 23.170127795527158, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 29009 + }, + { + "epoch": 23.170926517571885, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 29010 + }, + { + "epoch": 23.171725239616613, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 29011 + }, + { + "epoch": 23.17252396166134, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 29012 + }, + { + "epoch": 23.173322683706072, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 29013 + }, + { + "epoch": 23.1741214057508, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 29014 + }, + { + "epoch": 23.174920127795527, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29015 + }, + { + "epoch": 23.175718849840255, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 29016 + }, + { + "epoch": 23.176517571884983, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 29017 + }, + { + "epoch": 23.177316293929714, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 29018 + }, + { + "epoch": 23.17811501597444, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 29019 + }, + { + "epoch": 23.17891373801917, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 29020 + }, + { + "epoch": 23.179712460063897, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 29021 + }, + { + "epoch": 23.180511182108628, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 29022 + }, + { + "epoch": 23.181309904153355, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 29023 + }, + { + "epoch": 23.182108626198083, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29024 + }, + { + "epoch": 23.18290734824281, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 29025 + }, + { + "epoch": 23.18370607028754, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 29026 + }, + { + "epoch": 23.18450479233227, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 29027 + }, + { + "epoch": 23.185303514376997, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 29028 + }, + { + "epoch": 23.186102236421725, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 29029 + }, + { + "epoch": 23.186900958466452, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29030 + }, + { + "epoch": 23.187699680511184, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 29031 + }, + { + "epoch": 23.18849840255591, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 29032 + }, + { + "epoch": 23.18929712460064, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 29033 + }, + { + "epoch": 23.190095846645367, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29034 + }, + { + "epoch": 23.190894568690094, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 29035 + }, + { + "epoch": 23.191693290734825, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 29036 + }, + { + "epoch": 23.192492012779553, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 29037 + }, + { + "epoch": 23.19329073482428, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 29038 + }, + { + "epoch": 23.19408945686901, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 29039 + }, + { + "epoch": 23.19488817891374, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 29040 + }, + { + "epoch": 23.195686900958467, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 29041 + }, + { + "epoch": 23.196485623003195, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 29042 + }, + { + "epoch": 23.197284345047922, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 29043 + }, + { + "epoch": 23.19808306709265, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29044 + }, + { + "epoch": 23.19888178913738, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 29045 + }, + { + "epoch": 23.19968051118211, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 29046 + }, + { + "epoch": 23.200479233226837, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 29047 + }, + { + "epoch": 23.201277955271564, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 29048 + }, + { + "epoch": 23.202076677316295, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 29049 + }, + { + "epoch": 23.202875399361023, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 29050 + }, + { + "epoch": 23.20367412140575, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0762, + "step": 29051 + }, + { + "epoch": 23.20447284345048, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 29052 + }, + { + "epoch": 23.205271565495206, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 29053 + }, + { + "epoch": 23.206070287539937, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 29054 + }, + { + "epoch": 23.206869009584665, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29055 + }, + { + "epoch": 23.207667731629392, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 29056 + }, + { + "epoch": 23.20846645367412, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 29057 + }, + { + "epoch": 23.20926517571885, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29058 + }, + { + "epoch": 23.21006389776358, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 29059 + }, + { + "epoch": 23.210862619808307, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 29060 + }, + { + "epoch": 23.211661341853034, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 29061 + }, + { + "epoch": 23.212460063897762, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 29062 + }, + { + "epoch": 23.213258785942493, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 29063 + }, + { + "epoch": 23.21405750798722, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29064 + }, + { + "epoch": 23.21485623003195, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 29065 + }, + { + "epoch": 23.215654952076676, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 29066 + }, + { + "epoch": 23.216453674121407, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 29067 + }, + { + "epoch": 23.217252396166135, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 29068 + }, + { + "epoch": 23.218051118210862, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 29069 + }, + { + "epoch": 23.21884984025559, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 29070 + }, + { + "epoch": 23.21964856230032, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 29071 + }, + { + "epoch": 23.22044728434505, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 29072 + }, + { + "epoch": 23.221246006389777, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 29073 + }, + { + "epoch": 23.222044728434504, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 29074 + }, + { + "epoch": 23.222843450479232, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 29075 + }, + { + "epoch": 23.223642172523963, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 29076 + }, + { + "epoch": 23.22444089456869, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 29077 + }, + { + "epoch": 23.22523961661342, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 29078 + }, + { + "epoch": 23.226038338658146, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 29079 + }, + { + "epoch": 23.226837060702877, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 29080 + }, + { + "epoch": 23.227635782747605, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.076, + "step": 29081 + }, + { + "epoch": 23.228434504792332, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 29082 + }, + { + "epoch": 23.22923322683706, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 29083 + }, + { + "epoch": 23.230031948881788, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 29084 + }, + { + "epoch": 23.23083067092652, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 29085 + }, + { + "epoch": 23.231629392971247, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 29086 + }, + { + "epoch": 23.232428115015974, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 29087 + }, + { + "epoch": 23.233226837060702, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 29088 + }, + { + "epoch": 23.234025559105433, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 29089 + }, + { + "epoch": 23.23482428115016, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29090 + }, + { + "epoch": 23.23562300319489, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29091 + }, + { + "epoch": 23.236421725239616, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29092 + }, + { + "epoch": 23.237220447284344, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29093 + }, + { + "epoch": 23.238019169329075, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 29094 + }, + { + "epoch": 23.238817891373802, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 29095 + }, + { + "epoch": 23.23961661341853, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 29096 + }, + { + "epoch": 23.240415335463258, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29097 + }, + { + "epoch": 23.24121405750799, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 29098 + }, + { + "epoch": 23.242012779552716, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 29099 + }, + { + "epoch": 23.242811501597444, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 29100 + }, + { + "epoch": 23.24361022364217, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 29101 + }, + { + "epoch": 23.2444089456869, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 29102 + }, + { + "epoch": 23.24520766773163, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 29103 + }, + { + "epoch": 23.24600638977636, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 29104 + }, + { + "epoch": 23.246805111821086, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 29105 + }, + { + "epoch": 23.247603833865814, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 29106 + }, + { + "epoch": 23.248402555910545, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 29107 + }, + { + "epoch": 23.249201277955272, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 29108 + }, + { + "epoch": 23.25, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 29109 + }, + { + "epoch": 23.250798722044728, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 29110 + }, + { + "epoch": 23.251597444089455, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 29111 + }, + { + "epoch": 23.252396166134186, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 29112 + }, + { + "epoch": 23.253194888178914, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 29113 + }, + { + "epoch": 23.25399361022364, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 29114 + }, + { + "epoch": 23.25479233226837, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 29115 + }, + { + "epoch": 23.2555910543131, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 29116 + }, + { + "epoch": 23.25638977635783, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 29117 + }, + { + "epoch": 23.257188498402556, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 29118 + }, + { + "epoch": 23.257987220447284, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 29119 + }, + { + "epoch": 23.25878594249201, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 29120 + }, + { + "epoch": 23.259584664536742, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 29121 + }, + { + "epoch": 23.26038338658147, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29122 + }, + { + "epoch": 23.261182108626198, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 29123 + }, + { + "epoch": 23.261980830670925, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29124 + }, + { + "epoch": 23.262779552715656, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 29125 + }, + { + "epoch": 23.263578274760384, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 29126 + }, + { + "epoch": 23.26437699680511, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29127 + }, + { + "epoch": 23.26517571884984, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 29128 + }, + { + "epoch": 23.265974440894567, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 29129 + }, + { + "epoch": 23.266773162939298, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 29130 + }, + { + "epoch": 23.267571884984026, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 29131 + }, + { + "epoch": 23.268370607028753, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 29132 + }, + { + "epoch": 23.26916932907348, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29133 + }, + { + "epoch": 23.269968051118212, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 29134 + }, + { + "epoch": 23.27076677316294, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 29135 + }, + { + "epoch": 23.271565495207668, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 29136 + }, + { + "epoch": 23.272364217252395, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 29137 + }, + { + "epoch": 23.273162939297123, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 29138 + }, + { + "epoch": 23.273961661341854, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 29139 + }, + { + "epoch": 23.27476038338658, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 29140 + }, + { + "epoch": 23.27555910543131, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 29141 + }, + { + "epoch": 23.276357827476037, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 29142 + }, + { + "epoch": 23.277156549520768, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29143 + }, + { + "epoch": 23.277955271565496, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 29144 + }, + { + "epoch": 23.278753993610223, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 29145 + }, + { + "epoch": 23.27955271565495, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 29146 + }, + { + "epoch": 23.28035143769968, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 29147 + }, + { + "epoch": 23.28115015974441, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 29148 + }, + { + "epoch": 23.281948881789138, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 29149 + }, + { + "epoch": 23.282747603833865, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 29150 + }, + { + "epoch": 23.283546325878593, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 29151 + }, + { + "epoch": 23.284345047923324, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 29152 + }, + { + "epoch": 23.28514376996805, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 29153 + }, + { + "epoch": 23.28594249201278, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 29154 + }, + { + "epoch": 23.286741214057507, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 29155 + }, + { + "epoch": 23.287539936102238, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 29156 + }, + { + "epoch": 23.288338658146966, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29157 + }, + { + "epoch": 23.289137380191693, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 29158 + }, + { + "epoch": 23.28993610223642, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 29159 + }, + { + "epoch": 23.29073482428115, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 29160 + }, + { + "epoch": 23.29153354632588, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 29161 + }, + { + "epoch": 23.292332268370608, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29162 + }, + { + "epoch": 23.293130990415335, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 29163 + }, + { + "epoch": 23.293929712460063, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 29164 + }, + { + "epoch": 23.294728434504794, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0752, + "step": 29165 + }, + { + "epoch": 23.29552715654952, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0743, + "step": 29166 + }, + { + "epoch": 23.29632587859425, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 29167 + }, + { + "epoch": 23.297124600638977, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 29168 + }, + { + "epoch": 23.297923322683705, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 29169 + }, + { + "epoch": 23.298722044728436, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 29170 + }, + { + "epoch": 23.299520766773163, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29171 + }, + { + "epoch": 23.30031948881789, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29172 + }, + { + "epoch": 23.30111821086262, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 29173 + }, + { + "epoch": 23.30191693290735, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 29174 + }, + { + "epoch": 23.302715654952078, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 29175 + }, + { + "epoch": 23.303514376996805, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 29176 + }, + { + "epoch": 23.304313099041533, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 29177 + }, + { + "epoch": 23.30511182108626, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29178 + }, + { + "epoch": 23.30591054313099, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 29179 + }, + { + "epoch": 23.30670926517572, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 29180 + }, + { + "epoch": 23.307507987220447, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29181 + }, + { + "epoch": 23.308306709265175, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 29182 + }, + { + "epoch": 23.309105431309906, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 29183 + }, + { + "epoch": 23.309904153354633, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 29184 + }, + { + "epoch": 23.31070287539936, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 29185 + }, + { + "epoch": 23.31150159744409, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29186 + }, + { + "epoch": 23.312300319488816, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 29187 + }, + { + "epoch": 23.313099041533548, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29188 + }, + { + "epoch": 23.313897763578275, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 29189 + }, + { + "epoch": 23.314696485623003, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 29190 + }, + { + "epoch": 23.31549520766773, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 29191 + }, + { + "epoch": 23.31629392971246, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 29192 + }, + { + "epoch": 23.31709265175719, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29193 + }, + { + "epoch": 23.317891373801917, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 29194 + }, + { + "epoch": 23.318690095846645, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 29195 + }, + { + "epoch": 23.319488817891372, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 29196 + }, + { + "epoch": 23.320287539936103, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 29197 + }, + { + "epoch": 23.32108626198083, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 29198 + }, + { + "epoch": 23.32188498402556, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 29199 + }, + { + "epoch": 23.322683706070286, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 29200 + }, + { + "epoch": 23.323482428115017, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29201 + }, + { + "epoch": 23.324281150159745, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 29202 + }, + { + "epoch": 23.325079872204473, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 29203 + }, + { + "epoch": 23.3258785942492, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 29204 + }, + { + "epoch": 23.326677316293928, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 29205 + }, + { + "epoch": 23.32747603833866, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 29206 + }, + { + "epoch": 23.328274760383387, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29207 + }, + { + "epoch": 23.329073482428115, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 29208 + }, + { + "epoch": 23.329872204472842, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 29209 + }, + { + "epoch": 23.330670926517573, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 29210 + }, + { + "epoch": 23.3314696485623, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 29211 + }, + { + "epoch": 23.33226837060703, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 29212 + }, + { + "epoch": 23.333067092651756, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 29213 + }, + { + "epoch": 23.333865814696484, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 29214 + }, + { + "epoch": 23.334664536741215, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 29215 + }, + { + "epoch": 23.335463258785943, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 29216 + }, + { + "epoch": 23.33626198083067, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 29217 + }, + { + "epoch": 23.337060702875398, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 29218 + }, + { + "epoch": 23.33785942492013, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 29219 + }, + { + "epoch": 23.338658146964857, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 29220 + }, + { + "epoch": 23.339456869009584, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29221 + }, + { + "epoch": 23.340255591054312, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 29222 + }, + { + "epoch": 23.341054313099043, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 29223 + }, + { + "epoch": 23.34185303514377, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 29224 + }, + { + "epoch": 23.3426517571885, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 29225 + }, + { + "epoch": 23.343450479233226, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 29226 + }, + { + "epoch": 23.344249201277954, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 29227 + }, + { + "epoch": 23.345047923322685, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 29228 + }, + { + "epoch": 23.345846645367413, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 29229 + }, + { + "epoch": 23.34664536741214, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 29230 + }, + { + "epoch": 23.347444089456868, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29231 + }, + { + "epoch": 23.3482428115016, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 29232 + }, + { + "epoch": 23.349041533546327, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 29233 + }, + { + "epoch": 23.349840255591054, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 29234 + }, + { + "epoch": 23.350638977635782, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29235 + }, + { + "epoch": 23.35143769968051, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 29236 + }, + { + "epoch": 23.35223642172524, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 29237 + }, + { + "epoch": 23.35303514376997, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 29238 + }, + { + "epoch": 23.353833865814696, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 29239 + }, + { + "epoch": 23.354632587859424, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 29240 + }, + { + "epoch": 23.355431309904155, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29241 + }, + { + "epoch": 23.356230031948883, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29242 + }, + { + "epoch": 23.35702875399361, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 29243 + }, + { + "epoch": 23.357827476038338, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 29244 + }, + { + "epoch": 23.358626198083066, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 29245 + }, + { + "epoch": 23.359424920127797, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 29246 + }, + { + "epoch": 23.360223642172524, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 29247 + }, + { + "epoch": 23.361022364217252, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 29248 + }, + { + "epoch": 23.36182108626198, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 29249 + }, + { + "epoch": 23.36261980830671, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 29250 + }, + { + "epoch": 23.36341853035144, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 29251 + }, + { + "epoch": 23.364217252396166, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 29252 + }, + { + "epoch": 23.365015974440894, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 29253 + }, + { + "epoch": 23.36581469648562, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 29254 + }, + { + "epoch": 23.366613418530353, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 29255 + }, + { + "epoch": 23.36741214057508, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29256 + }, + { + "epoch": 23.368210862619808, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 29257 + }, + { + "epoch": 23.369009584664536, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 29258 + }, + { + "epoch": 23.369808306709267, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29259 + }, + { + "epoch": 23.370607028753994, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 29260 + }, + { + "epoch": 23.371405750798722, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29261 + }, + { + "epoch": 23.37220447284345, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 29262 + }, + { + "epoch": 23.373003194888177, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 29263 + }, + { + "epoch": 23.37380191693291, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 29264 + }, + { + "epoch": 23.374600638977636, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 29265 + }, + { + "epoch": 23.375399361022364, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 29266 + }, + { + "epoch": 23.37619808306709, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 29267 + }, + { + "epoch": 23.376996805111823, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 29268 + }, + { + "epoch": 23.37779552715655, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 29269 + }, + { + "epoch": 23.378594249201278, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 29270 + }, + { + "epoch": 23.379392971246006, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 29271 + }, + { + "epoch": 23.380191693290733, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29272 + }, + { + "epoch": 23.380990415335464, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 29273 + }, + { + "epoch": 23.381789137380192, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29274 + }, + { + "epoch": 23.38258785942492, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 29275 + }, + { + "epoch": 23.383386581469647, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 29276 + }, + { + "epoch": 23.38418530351438, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 29277 + }, + { + "epoch": 23.384984025559106, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 29278 + }, + { + "epoch": 23.385782747603834, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29279 + }, + { + "epoch": 23.38658146964856, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 29280 + }, + { + "epoch": 23.38738019169329, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 29281 + }, + { + "epoch": 23.38817891373802, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 29282 + }, + { + "epoch": 23.388977635782748, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 29283 + }, + { + "epoch": 23.389776357827476, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 29284 + }, + { + "epoch": 23.390575079872203, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 29285 + }, + { + "epoch": 23.391373801916934, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29286 + }, + { + "epoch": 23.392172523961662, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 29287 + }, + { + "epoch": 23.39297124600639, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 29288 + }, + { + "epoch": 23.393769968051117, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 29289 + }, + { + "epoch": 23.394568690095845, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29290 + }, + { + "epoch": 23.395367412140576, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 29291 + }, + { + "epoch": 23.396166134185304, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 29292 + }, + { + "epoch": 23.39696485623003, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 29293 + }, + { + "epoch": 23.39776357827476, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 29294 + }, + { + "epoch": 23.39856230031949, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 29295 + }, + { + "epoch": 23.399361022364218, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 29296 + }, + { + "epoch": 23.400159744408946, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29297 + }, + { + "epoch": 23.400958466453673, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 29298 + }, + { + "epoch": 23.4017571884984, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 29299 + }, + { + "epoch": 23.402555910543132, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 29300 + }, + { + "epoch": 23.40335463258786, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 29301 + }, + { + "epoch": 23.404153354632587, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29302 + }, + { + "epoch": 23.404952076677315, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 29303 + }, + { + "epoch": 23.405750798722046, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 29304 + }, + { + "epoch": 23.406549520766774, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 29305 + }, + { + "epoch": 23.4073482428115, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 29306 + }, + { + "epoch": 23.40814696485623, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 29307 + }, + { + "epoch": 23.408945686900957, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 29308 + }, + { + "epoch": 23.409744408945688, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29309 + }, + { + "epoch": 23.410543130990416, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29310 + }, + { + "epoch": 23.411341853035143, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 29311 + }, + { + "epoch": 23.41214057507987, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 29312 + }, + { + "epoch": 23.412939297124602, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 29313 + }, + { + "epoch": 23.41373801916933, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 29314 + }, + { + "epoch": 23.414536741214057, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 29315 + }, + { + "epoch": 23.415335463258785, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 29316 + }, + { + "epoch": 23.416134185303516, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 29317 + }, + { + "epoch": 23.416932907348244, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 29318 + }, + { + "epoch": 23.41773162939297, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 29319 + }, + { + "epoch": 23.4185303514377, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 29320 + }, + { + "epoch": 23.419329073482427, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 29321 + }, + { + "epoch": 23.420127795527158, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 29322 + }, + { + "epoch": 23.420926517571885, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 29323 + }, + { + "epoch": 23.421725239616613, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 29324 + }, + { + "epoch": 23.42252396166134, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 29325 + }, + { + "epoch": 23.423322683706072, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 29326 + }, + { + "epoch": 23.4241214057508, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 29327 + }, + { + "epoch": 23.424920127795527, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 29328 + }, + { + "epoch": 23.425718849840255, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29329 + }, + { + "epoch": 23.426517571884983, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 29330 + }, + { + "epoch": 23.427316293929714, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 29331 + }, + { + "epoch": 23.42811501597444, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 29332 + }, + { + "epoch": 23.42891373801917, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29333 + }, + { + "epoch": 23.429712460063897, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 29334 + }, + { + "epoch": 23.430511182108628, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 29335 + }, + { + "epoch": 23.431309904153355, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 29336 + }, + { + "epoch": 23.432108626198083, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 29337 + }, + { + "epoch": 23.43290734824281, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 29338 + }, + { + "epoch": 23.43370607028754, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 29339 + }, + { + "epoch": 23.43450479233227, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 29340 + }, + { + "epoch": 23.435303514376997, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 29341 + }, + { + "epoch": 23.436102236421725, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 29342 + }, + { + "epoch": 23.436900958466452, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 29343 + }, + { + "epoch": 23.437699680511184, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 29344 + }, + { + "epoch": 23.43849840255591, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 29345 + }, + { + "epoch": 23.43929712460064, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 29346 + }, + { + "epoch": 23.440095846645367, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 29347 + }, + { + "epoch": 23.440894568690094, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 29348 + }, + { + "epoch": 23.441693290734825, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 29349 + }, + { + "epoch": 23.442492012779553, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 29350 + }, + { + "epoch": 23.44329073482428, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 29351 + }, + { + "epoch": 23.44408945686901, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 29352 + }, + { + "epoch": 23.44488817891374, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 29353 + }, + { + "epoch": 23.445686900958467, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 29354 + }, + { + "epoch": 23.446485623003195, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 29355 + }, + { + "epoch": 23.447284345047922, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 29356 + }, + { + "epoch": 23.44808306709265, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 29357 + }, + { + "epoch": 23.44888178913738, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 29358 + }, + { + "epoch": 23.44968051118211, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 29359 + }, + { + "epoch": 23.450479233226837, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 29360 + }, + { + "epoch": 23.451277955271564, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0756, + "step": 29361 + }, + { + "epoch": 23.452076677316295, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 29362 + }, + { + "epoch": 23.452875399361023, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 29363 + }, + { + "epoch": 23.45367412140575, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 29364 + }, + { + "epoch": 23.45447284345048, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 29365 + }, + { + "epoch": 23.455271565495206, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 29366 + }, + { + "epoch": 23.456070287539937, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 29367 + }, + { + "epoch": 23.456869009584665, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 29368 + }, + { + "epoch": 23.457667731629392, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 29369 + }, + { + "epoch": 23.45846645367412, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 29370 + }, + { + "epoch": 23.45926517571885, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 29371 + }, + { + "epoch": 23.46006389776358, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 29372 + }, + { + "epoch": 23.460862619808307, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 29373 + }, + { + "epoch": 23.461661341853034, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29374 + }, + { + "epoch": 23.462460063897762, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 29375 + }, + { + "epoch": 23.463258785942493, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 29376 + }, + { + "epoch": 23.46405750798722, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 29377 + }, + { + "epoch": 23.46485623003195, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 29378 + }, + { + "epoch": 23.465654952076676, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 29379 + }, + { + "epoch": 23.466453674121407, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 29380 + }, + { + "epoch": 23.467252396166135, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 29381 + }, + { + "epoch": 23.468051118210862, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 29382 + }, + { + "epoch": 23.46884984025559, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 29383 + }, + { + "epoch": 23.46964856230032, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 29384 + }, + { + "epoch": 23.47044728434505, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 29385 + }, + { + "epoch": 23.471246006389777, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 29386 + }, + { + "epoch": 23.472044728434504, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 29387 + }, + { + "epoch": 23.472843450479232, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 29388 + }, + { + "epoch": 23.473642172523963, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29389 + }, + { + "epoch": 23.47444089456869, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 29390 + }, + { + "epoch": 23.47523961661342, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 29391 + }, + { + "epoch": 23.476038338658146, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 29392 + }, + { + "epoch": 23.476837060702877, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 29393 + }, + { + "epoch": 23.477635782747605, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 29394 + }, + { + "epoch": 23.478434504792332, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 29395 + }, + { + "epoch": 23.47923322683706, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29396 + }, + { + "epoch": 23.480031948881788, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 29397 + }, + { + "epoch": 23.48083067092652, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 29398 + }, + { + "epoch": 23.481629392971247, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 29399 + }, + { + "epoch": 23.482428115015974, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29400 + }, + { + "epoch": 23.483226837060702, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 29401 + }, + { + "epoch": 23.484025559105433, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 29402 + }, + { + "epoch": 23.48482428115016, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 29403 + }, + { + "epoch": 23.48562300319489, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 29404 + }, + { + "epoch": 23.486421725239616, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 29405 + }, + { + "epoch": 23.487220447284344, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 29406 + }, + { + "epoch": 23.488019169329075, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 29407 + }, + { + "epoch": 23.488817891373802, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 29408 + }, + { + "epoch": 23.48961661341853, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 29409 + }, + { + "epoch": 23.490415335463258, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 29410 + }, + { + "epoch": 23.49121405750799, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 29411 + }, + { + "epoch": 23.492012779552716, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 29412 + }, + { + "epoch": 23.492811501597444, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 29413 + }, + { + "epoch": 23.49361022364217, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 29414 + }, + { + "epoch": 23.4944089456869, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 29415 + }, + { + "epoch": 23.49520766773163, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 29416 + }, + { + "epoch": 23.49600638977636, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 29417 + }, + { + "epoch": 23.496805111821086, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 29418 + }, + { + "epoch": 23.497603833865814, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29419 + }, + { + "epoch": 23.498402555910545, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 29420 + }, + { + "epoch": 23.499201277955272, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 29421 + }, + { + "epoch": 23.5, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 29422 + }, + { + "epoch": 23.500798722044728, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 29423 + }, + { + "epoch": 23.501597444089455, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 29424 + }, + { + "epoch": 23.502396166134186, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 29425 + }, + { + "epoch": 23.503194888178914, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 29426 + }, + { + "epoch": 23.50399361022364, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 29427 + }, + { + "epoch": 23.50479233226837, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 29428 + }, + { + "epoch": 23.5055910543131, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 29429 + }, + { + "epoch": 23.50638977635783, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 29430 + }, + { + "epoch": 23.507188498402556, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 29431 + }, + { + "epoch": 23.507987220447284, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 29432 + }, + { + "epoch": 23.50878594249201, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 29433 + }, + { + "epoch": 23.509584664536742, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29434 + }, + { + "epoch": 23.51038338658147, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 29435 + }, + { + "epoch": 23.511182108626198, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 29436 + }, + { + "epoch": 23.511980830670925, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 29437 + }, + { + "epoch": 23.512779552715656, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29438 + }, + { + "epoch": 23.513578274760384, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 29439 + }, + { + "epoch": 23.51437699680511, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 29440 + }, + { + "epoch": 23.51517571884984, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 29441 + }, + { + "epoch": 23.515974440894567, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 29442 + }, + { + "epoch": 23.516773162939298, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 29443 + }, + { + "epoch": 23.517571884984026, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 29444 + }, + { + "epoch": 23.518370607028753, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 29445 + }, + { + "epoch": 23.51916932907348, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 29446 + }, + { + "epoch": 23.519968051118212, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 29447 + }, + { + "epoch": 23.52076677316294, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 29448 + }, + { + "epoch": 23.521565495207668, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 29449 + }, + { + "epoch": 23.522364217252395, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29450 + }, + { + "epoch": 23.523162939297123, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 29451 + }, + { + "epoch": 23.523961661341854, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 29452 + }, + { + "epoch": 23.52476038338658, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 29453 + }, + { + "epoch": 23.52555910543131, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 29454 + }, + { + "epoch": 23.526357827476037, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 29455 + }, + { + "epoch": 23.527156549520768, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 29456 + }, + { + "epoch": 23.527955271565496, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 29457 + }, + { + "epoch": 23.528753993610223, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 29458 + }, + { + "epoch": 23.52955271565495, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 29459 + }, + { + "epoch": 23.53035143769968, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 29460 + }, + { + "epoch": 23.53115015974441, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 29461 + }, + { + "epoch": 23.531948881789138, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 29462 + }, + { + "epoch": 23.532747603833865, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 29463 + }, + { + "epoch": 23.533546325878593, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 29464 + }, + { + "epoch": 23.534345047923324, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 29465 + }, + { + "epoch": 23.53514376996805, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 29466 + }, + { + "epoch": 23.53594249201278, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 29467 + }, + { + "epoch": 23.536741214057507, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 29468 + }, + { + "epoch": 23.537539936102235, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 29469 + }, + { + "epoch": 23.538338658146966, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 29470 + }, + { + "epoch": 23.539137380191693, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 29471 + }, + { + "epoch": 23.53993610223642, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 29472 + }, + { + "epoch": 23.54073482428115, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29473 + }, + { + "epoch": 23.54153354632588, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29474 + }, + { + "epoch": 23.542332268370608, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 29475 + }, + { + "epoch": 23.543130990415335, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 29476 + }, + { + "epoch": 23.543929712460063, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 29477 + }, + { + "epoch": 23.544728434504794, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 29478 + }, + { + "epoch": 23.54552715654952, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 29479 + }, + { + "epoch": 23.54632587859425, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 29480 + }, + { + "epoch": 23.547124600638977, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 29481 + }, + { + "epoch": 23.547923322683705, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 29482 + }, + { + "epoch": 23.548722044728436, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 29483 + }, + { + "epoch": 23.549520766773163, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 29484 + }, + { + "epoch": 23.55031948881789, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 29485 + }, + { + "epoch": 23.55111821086262, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 29486 + }, + { + "epoch": 23.55191693290735, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 29487 + }, + { + "epoch": 23.552715654952078, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 29488 + }, + { + "epoch": 23.553514376996805, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29489 + }, + { + "epoch": 23.554313099041533, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 29490 + }, + { + "epoch": 23.55511182108626, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 29491 + }, + { + "epoch": 23.55591054313099, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 29492 + }, + { + "epoch": 23.55670926517572, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 29493 + }, + { + "epoch": 23.557507987220447, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 29494 + }, + { + "epoch": 23.558306709265175, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29495 + }, + { + "epoch": 23.559105431309906, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 29496 + }, + { + "epoch": 23.559904153354633, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 29497 + }, + { + "epoch": 23.56070287539936, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29498 + }, + { + "epoch": 23.56150159744409, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 29499 + }, + { + "epoch": 23.562300319488816, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29500 + }, + { + "epoch": 23.563099041533548, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29501 + }, + { + "epoch": 23.563897763578275, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 29502 + }, + { + "epoch": 23.564696485623003, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 29503 + }, + { + "epoch": 23.56549520766773, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 29504 + }, + { + "epoch": 23.56629392971246, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 29505 + }, + { + "epoch": 23.56709265175719, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 29506 + }, + { + "epoch": 23.567891373801917, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 29507 + }, + { + "epoch": 23.568690095846645, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 29508 + }, + { + "epoch": 23.569488817891372, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 29509 + }, + { + "epoch": 23.570287539936103, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29510 + }, + { + "epoch": 23.57108626198083, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29511 + }, + { + "epoch": 23.57188498402556, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 29512 + }, + { + "epoch": 23.572683706070286, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 29513 + }, + { + "epoch": 23.573482428115017, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 29514 + }, + { + "epoch": 23.574281150159745, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 29515 + }, + { + "epoch": 23.575079872204473, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 29516 + }, + { + "epoch": 23.5758785942492, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 29517 + }, + { + "epoch": 23.576677316293928, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 29518 + }, + { + "epoch": 23.57747603833866, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 29519 + }, + { + "epoch": 23.578274760383387, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 29520 + }, + { + "epoch": 23.579073482428115, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 29521 + }, + { + "epoch": 23.579872204472842, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 29522 + }, + { + "epoch": 23.580670926517573, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 29523 + }, + { + "epoch": 23.5814696485623, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 29524 + }, + { + "epoch": 23.58226837060703, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 29525 + }, + { + "epoch": 23.583067092651756, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 29526 + }, + { + "epoch": 23.583865814696484, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 29527 + }, + { + "epoch": 23.584664536741215, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 29528 + }, + { + "epoch": 23.585463258785943, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 29529 + }, + { + "epoch": 23.58626198083067, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 29530 + }, + { + "epoch": 23.587060702875398, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29531 + }, + { + "epoch": 23.58785942492013, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29532 + }, + { + "epoch": 23.588658146964857, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 29533 + }, + { + "epoch": 23.589456869009584, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 29534 + }, + { + "epoch": 23.590255591054312, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 29535 + }, + { + "epoch": 23.591054313099043, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29536 + }, + { + "epoch": 23.59185303514377, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29537 + }, + { + "epoch": 23.5926517571885, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 29538 + }, + { + "epoch": 23.593450479233226, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 29539 + }, + { + "epoch": 23.594249201277954, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 29540 + }, + { + "epoch": 23.595047923322685, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29541 + }, + { + "epoch": 23.595846645367413, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 29542 + }, + { + "epoch": 23.59664536741214, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 29543 + }, + { + "epoch": 23.597444089456868, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29544 + }, + { + "epoch": 23.5982428115016, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 29545 + }, + { + "epoch": 23.599041533546327, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 29546 + }, + { + "epoch": 23.599840255591054, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 29547 + }, + { + "epoch": 23.600638977635782, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 29548 + }, + { + "epoch": 23.60143769968051, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 29549 + }, + { + "epoch": 23.60223642172524, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 29550 + }, + { + "epoch": 23.60303514376997, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 29551 + }, + { + "epoch": 23.603833865814696, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 29552 + }, + { + "epoch": 23.604632587859424, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 29553 + }, + { + "epoch": 23.605431309904155, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 29554 + }, + { + "epoch": 23.606230031948883, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 29555 + }, + { + "epoch": 23.60702875399361, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 29556 + }, + { + "epoch": 23.607827476038338, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 29557 + }, + { + "epoch": 23.608626198083066, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 29558 + }, + { + "epoch": 23.609424920127797, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 29559 + }, + { + "epoch": 23.610223642172524, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 29560 + }, + { + "epoch": 23.611022364217252, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 29561 + }, + { + "epoch": 23.61182108626198, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 29562 + }, + { + "epoch": 23.61261980830671, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 29563 + }, + { + "epoch": 23.61341853035144, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 29564 + }, + { + "epoch": 23.614217252396166, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 29565 + }, + { + "epoch": 23.615015974440894, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 29566 + }, + { + "epoch": 23.61581469648562, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 29567 + }, + { + "epoch": 23.616613418530353, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 29568 + }, + { + "epoch": 23.61741214057508, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29569 + }, + { + "epoch": 23.618210862619808, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 29570 + }, + { + "epoch": 23.619009584664536, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 29571 + }, + { + "epoch": 23.619808306709267, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 29572 + }, + { + "epoch": 23.620607028753994, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29573 + }, + { + "epoch": 23.621405750798722, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 29574 + }, + { + "epoch": 23.62220447284345, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 29575 + }, + { + "epoch": 23.623003194888177, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 29576 + }, + { + "epoch": 23.62380191693291, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 29577 + }, + { + "epoch": 23.624600638977636, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 29578 + }, + { + "epoch": 23.625399361022364, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29579 + }, + { + "epoch": 23.62619808306709, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 29580 + }, + { + "epoch": 23.626996805111823, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 29581 + }, + { + "epoch": 23.62779552715655, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 29582 + }, + { + "epoch": 23.628594249201278, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 29583 + }, + { + "epoch": 23.629392971246006, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29584 + }, + { + "epoch": 23.630191693290733, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 29585 + }, + { + "epoch": 23.630990415335464, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 29586 + }, + { + "epoch": 23.631789137380192, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 29587 + }, + { + "epoch": 23.63258785942492, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29588 + }, + { + "epoch": 23.633386581469647, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 29589 + }, + { + "epoch": 23.63418530351438, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 29590 + }, + { + "epoch": 23.634984025559106, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 29591 + }, + { + "epoch": 23.635782747603834, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 29592 + }, + { + "epoch": 23.63658146964856, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29593 + }, + { + "epoch": 23.63738019169329, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29594 + }, + { + "epoch": 23.63817891373802, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 29595 + }, + { + "epoch": 23.638977635782748, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 29596 + }, + { + "epoch": 23.639776357827476, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 29597 + }, + { + "epoch": 23.640575079872203, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29598 + }, + { + "epoch": 23.641373801916934, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 29599 + }, + { + "epoch": 23.642172523961662, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 29600 + }, + { + "epoch": 23.64297124600639, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 29601 + }, + { + "epoch": 23.643769968051117, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 29602 + }, + { + "epoch": 23.644568690095845, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 29603 + }, + { + "epoch": 23.645367412140576, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29604 + }, + { + "epoch": 23.646166134185304, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 29605 + }, + { + "epoch": 23.64696485623003, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 29606 + }, + { + "epoch": 23.64776357827476, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 29607 + }, + { + "epoch": 23.64856230031949, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 29608 + }, + { + "epoch": 23.649361022364218, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 29609 + }, + { + "epoch": 23.650159744408946, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 29610 + }, + { + "epoch": 23.650958466453673, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 29611 + }, + { + "epoch": 23.6517571884984, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 29612 + }, + { + "epoch": 23.652555910543132, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 29613 + }, + { + "epoch": 23.65335463258786, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 29614 + }, + { + "epoch": 23.654153354632587, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 29615 + }, + { + "epoch": 23.654952076677315, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 29616 + }, + { + "epoch": 23.655750798722046, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 29617 + }, + { + "epoch": 23.656549520766774, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 29618 + }, + { + "epoch": 23.6573482428115, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29619 + }, + { + "epoch": 23.65814696485623, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29620 + }, + { + "epoch": 23.658945686900957, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 29621 + }, + { + "epoch": 23.659744408945688, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 29622 + }, + { + "epoch": 23.660543130990416, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 29623 + }, + { + "epoch": 23.661341853035143, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 29624 + }, + { + "epoch": 23.66214057507987, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 29625 + }, + { + "epoch": 23.662939297124602, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 29626 + }, + { + "epoch": 23.66373801916933, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 29627 + }, + { + "epoch": 23.664536741214057, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 29628 + }, + { + "epoch": 23.665335463258785, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 29629 + }, + { + "epoch": 23.666134185303516, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 29630 + }, + { + "epoch": 23.666932907348244, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 29631 + }, + { + "epoch": 23.66773162939297, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 29632 + }, + { + "epoch": 23.6685303514377, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29633 + }, + { + "epoch": 23.669329073482427, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 29634 + }, + { + "epoch": 23.670127795527158, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 29635 + }, + { + "epoch": 23.670926517571885, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 29636 + }, + { + "epoch": 23.671725239616613, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29637 + }, + { + "epoch": 23.67252396166134, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 29638 + }, + { + "epoch": 23.673322683706072, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 29639 + }, + { + "epoch": 23.6741214057508, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29640 + }, + { + "epoch": 23.674920127795527, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 29641 + }, + { + "epoch": 23.675718849840255, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 29642 + }, + { + "epoch": 23.676517571884983, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 29643 + }, + { + "epoch": 23.677316293929714, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29644 + }, + { + "epoch": 23.67811501597444, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 29645 + }, + { + "epoch": 23.67891373801917, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 29646 + }, + { + "epoch": 23.679712460063897, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 29647 + }, + { + "epoch": 23.680511182108628, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 29648 + }, + { + "epoch": 23.681309904153355, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 29649 + }, + { + "epoch": 23.682108626198083, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 29650 + }, + { + "epoch": 23.68290734824281, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 29651 + }, + { + "epoch": 23.68370607028754, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 29652 + }, + { + "epoch": 23.68450479233227, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 29653 + }, + { + "epoch": 23.685303514376997, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29654 + }, + { + "epoch": 23.686102236421725, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 29655 + }, + { + "epoch": 23.686900958466452, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 29656 + }, + { + "epoch": 23.687699680511184, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 29657 + }, + { + "epoch": 23.68849840255591, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0762, + "step": 29658 + }, + { + "epoch": 23.68929712460064, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 29659 + }, + { + "epoch": 23.690095846645367, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29660 + }, + { + "epoch": 23.690894568690094, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 29661 + }, + { + "epoch": 23.691693290734825, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 29662 + }, + { + "epoch": 23.692492012779553, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 29663 + }, + { + "epoch": 23.69329073482428, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 29664 + }, + { + "epoch": 23.69408945686901, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 29665 + }, + { + "epoch": 23.69488817891374, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29666 + }, + { + "epoch": 23.695686900958467, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29667 + }, + { + "epoch": 23.696485623003195, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 29668 + }, + { + "epoch": 23.697284345047922, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29669 + }, + { + "epoch": 23.69808306709265, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 29670 + }, + { + "epoch": 23.69888178913738, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 29671 + }, + { + "epoch": 23.69968051118211, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 29672 + }, + { + "epoch": 23.700479233226837, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 29673 + }, + { + "epoch": 23.701277955271564, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 29674 + }, + { + "epoch": 23.702076677316295, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29675 + }, + { + "epoch": 23.702875399361023, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 29676 + }, + { + "epoch": 23.70367412140575, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 29677 + }, + { + "epoch": 23.70447284345048, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 29678 + }, + { + "epoch": 23.705271565495206, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 29679 + }, + { + "epoch": 23.706070287539937, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 29680 + }, + { + "epoch": 23.706869009584665, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 29681 + }, + { + "epoch": 23.707667731629392, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.075, + "step": 29682 + }, + { + "epoch": 23.70846645367412, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 29683 + }, + { + "epoch": 23.70926517571885, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 29684 + }, + { + "epoch": 23.71006389776358, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 29685 + }, + { + "epoch": 23.710862619808307, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 29686 + }, + { + "epoch": 23.711661341853034, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 29687 + }, + { + "epoch": 23.712460063897765, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 29688 + }, + { + "epoch": 23.713258785942493, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 29689 + }, + { + "epoch": 23.71405750798722, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 29690 + }, + { + "epoch": 23.71485623003195, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 29691 + }, + { + "epoch": 23.715654952076676, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 29692 + }, + { + "epoch": 23.716453674121407, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 29693 + }, + { + "epoch": 23.717252396166135, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 29694 + }, + { + "epoch": 23.718051118210862, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 29695 + }, + { + "epoch": 23.71884984025559, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 29696 + }, + { + "epoch": 23.71964856230032, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 29697 + }, + { + "epoch": 23.72044728434505, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 29698 + }, + { + "epoch": 23.721246006389777, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 29699 + }, + { + "epoch": 23.722044728434504, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 29700 + }, + { + "epoch": 23.722843450479232, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 29701 + }, + { + "epoch": 23.723642172523963, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 29702 + }, + { + "epoch": 23.72444089456869, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 29703 + }, + { + "epoch": 23.72523961661342, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 29704 + }, + { + "epoch": 23.726038338658146, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 29705 + }, + { + "epoch": 23.726837060702877, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 29706 + }, + { + "epoch": 23.727635782747605, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 29707 + }, + { + "epoch": 23.728434504792332, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 29708 + }, + { + "epoch": 23.72923322683706, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 29709 + }, + { + "epoch": 23.730031948881788, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 29710 + }, + { + "epoch": 23.73083067092652, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 29711 + }, + { + "epoch": 23.731629392971247, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 29712 + }, + { + "epoch": 23.732428115015974, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 29713 + }, + { + "epoch": 23.733226837060702, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 29714 + }, + { + "epoch": 23.734025559105433, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 29715 + }, + { + "epoch": 23.73482428115016, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 29716 + }, + { + "epoch": 23.73562300319489, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 29717 + }, + { + "epoch": 23.736421725239616, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 29718 + }, + { + "epoch": 23.737220447284344, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 29719 + }, + { + "epoch": 23.738019169329075, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 29720 + }, + { + "epoch": 23.738817891373802, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 29721 + }, + { + "epoch": 23.73961661341853, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 29722 + }, + { + "epoch": 23.740415335463258, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 29723 + }, + { + "epoch": 23.74121405750799, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 29724 + }, + { + "epoch": 23.742012779552716, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 29725 + }, + { + "epoch": 23.742811501597444, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29726 + }, + { + "epoch": 23.74361022364217, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 29727 + }, + { + "epoch": 23.7444089456869, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 29728 + }, + { + "epoch": 23.74520766773163, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 29729 + }, + { + "epoch": 23.74600638977636, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29730 + }, + { + "epoch": 23.746805111821086, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29731 + }, + { + "epoch": 23.747603833865814, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 29732 + }, + { + "epoch": 23.748402555910545, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 29733 + }, + { + "epoch": 23.749201277955272, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 29734 + }, + { + "epoch": 23.75, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 29735 + }, + { + "epoch": 23.750798722044728, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 29736 + }, + { + "epoch": 23.751597444089455, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 29737 + }, + { + "epoch": 23.752396166134186, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 29738 + }, + { + "epoch": 23.753194888178914, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 29739 + }, + { + "epoch": 23.75399361022364, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29740 + }, + { + "epoch": 23.75479233226837, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 29741 + }, + { + "epoch": 23.7555910543131, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 29742 + }, + { + "epoch": 23.75638977635783, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 29743 + }, + { + "epoch": 23.757188498402556, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 29744 + }, + { + "epoch": 23.757987220447284, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29745 + }, + { + "epoch": 23.75878594249201, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29746 + }, + { + "epoch": 23.759584664536742, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 29747 + }, + { + "epoch": 23.76038338658147, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 29748 + }, + { + "epoch": 23.761182108626198, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 29749 + }, + { + "epoch": 23.761980830670925, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 29750 + }, + { + "epoch": 23.762779552715656, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 29751 + }, + { + "epoch": 23.763578274760384, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 29752 + }, + { + "epoch": 23.76437699680511, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 29753 + }, + { + "epoch": 23.76517571884984, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 29754 + }, + { + "epoch": 23.765974440894567, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 29755 + }, + { + "epoch": 23.766773162939298, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 29756 + }, + { + "epoch": 23.767571884984026, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 29757 + }, + { + "epoch": 23.768370607028753, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 29758 + }, + { + "epoch": 23.76916932907348, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0742, + "step": 29759 + }, + { + "epoch": 23.769968051118212, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29760 + }, + { + "epoch": 23.77076677316294, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 29761 + }, + { + "epoch": 23.771565495207668, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 29762 + }, + { + "epoch": 23.772364217252395, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 29763 + }, + { + "epoch": 23.773162939297123, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 29764 + }, + { + "epoch": 23.773961661341854, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 29765 + }, + { + "epoch": 23.77476038338658, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 29766 + }, + { + "epoch": 23.77555910543131, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 29767 + }, + { + "epoch": 23.776357827476037, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 29768 + }, + { + "epoch": 23.777156549520768, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 29769 + }, + { + "epoch": 23.777955271565496, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 29770 + }, + { + "epoch": 23.778753993610223, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 29771 + }, + { + "epoch": 23.77955271565495, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 29772 + }, + { + "epoch": 23.78035143769968, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29773 + }, + { + "epoch": 23.78115015974441, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 29774 + }, + { + "epoch": 23.781948881789138, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 29775 + }, + { + "epoch": 23.782747603833865, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 29776 + }, + { + "epoch": 23.783546325878593, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 29777 + }, + { + "epoch": 23.784345047923324, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 29778 + }, + { + "epoch": 23.78514376996805, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 29779 + }, + { + "epoch": 23.78594249201278, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 29780 + }, + { + "epoch": 23.786741214057507, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 29781 + }, + { + "epoch": 23.787539936102235, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 29782 + }, + { + "epoch": 23.788338658146966, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 29783 + }, + { + "epoch": 23.789137380191693, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 29784 + }, + { + "epoch": 23.78993610223642, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 29785 + }, + { + "epoch": 23.79073482428115, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 29786 + }, + { + "epoch": 23.79153354632588, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 29787 + }, + { + "epoch": 23.792332268370608, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 29788 + }, + { + "epoch": 23.793130990415335, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 29789 + }, + { + "epoch": 23.793929712460063, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 29790 + }, + { + "epoch": 23.794728434504794, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 29791 + }, + { + "epoch": 23.79552715654952, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 29792 + }, + { + "epoch": 23.79632587859425, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 29793 + }, + { + "epoch": 23.797124600638977, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 29794 + }, + { + "epoch": 23.797923322683705, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 29795 + }, + { + "epoch": 23.798722044728436, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 29796 + }, + { + "epoch": 23.799520766773163, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 29797 + }, + { + "epoch": 23.80031948881789, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 29798 + }, + { + "epoch": 23.80111821086262, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 29799 + }, + { + "epoch": 23.80191693290735, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 29800 + }, + { + "epoch": 23.802715654952078, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 29801 + }, + { + "epoch": 23.803514376996805, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 29802 + }, + { + "epoch": 23.804313099041533, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 29803 + }, + { + "epoch": 23.80511182108626, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 29804 + }, + { + "epoch": 23.80591054313099, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 29805 + }, + { + "epoch": 23.80670926517572, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 29806 + }, + { + "epoch": 23.807507987220447, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 29807 + }, + { + "epoch": 23.808306709265175, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 29808 + }, + { + "epoch": 23.809105431309906, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 29809 + }, + { + "epoch": 23.809904153354633, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 29810 + }, + { + "epoch": 23.81070287539936, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 29811 + }, + { + "epoch": 23.81150159744409, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 29812 + }, + { + "epoch": 23.812300319488816, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 29813 + }, + { + "epoch": 23.813099041533548, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29814 + }, + { + "epoch": 23.813897763578275, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 29815 + }, + { + "epoch": 23.814696485623003, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 29816 + }, + { + "epoch": 23.81549520766773, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0721, + "step": 29817 + }, + { + "epoch": 23.81629392971246, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 29818 + }, + { + "epoch": 23.81709265175719, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 29819 + }, + { + "epoch": 23.817891373801917, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29820 + }, + { + "epoch": 23.818690095846645, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 29821 + }, + { + "epoch": 23.819488817891372, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 29822 + }, + { + "epoch": 23.820287539936103, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 29823 + }, + { + "epoch": 23.82108626198083, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 29824 + }, + { + "epoch": 23.82188498402556, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 29825 + }, + { + "epoch": 23.822683706070286, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 29826 + }, + { + "epoch": 23.823482428115017, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 29827 + }, + { + "epoch": 23.824281150159745, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 29828 + }, + { + "epoch": 23.825079872204473, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 29829 + }, + { + "epoch": 23.8258785942492, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 29830 + }, + { + "epoch": 23.826677316293928, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 29831 + }, + { + "epoch": 23.82747603833866, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 29832 + }, + { + "epoch": 23.828274760383387, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 29833 + }, + { + "epoch": 23.829073482428115, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 29834 + }, + { + "epoch": 23.829872204472842, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 29835 + }, + { + "epoch": 23.830670926517573, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 29836 + }, + { + "epoch": 23.8314696485623, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 29837 + }, + { + "epoch": 23.83226837060703, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 29838 + }, + { + "epoch": 23.833067092651756, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 29839 + }, + { + "epoch": 23.833865814696484, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 29840 + }, + { + "epoch": 23.834664536741215, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 29841 + }, + { + "epoch": 23.835463258785943, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 29842 + }, + { + "epoch": 23.83626198083067, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 29843 + }, + { + "epoch": 23.837060702875398, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 29844 + }, + { + "epoch": 23.83785942492013, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 29845 + }, + { + "epoch": 23.838658146964857, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 29846 + }, + { + "epoch": 23.839456869009584, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 29847 + }, + { + "epoch": 23.840255591054312, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 29848 + }, + { + "epoch": 23.841054313099043, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 29849 + }, + { + "epoch": 23.84185303514377, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 29850 + }, + { + "epoch": 23.8426517571885, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 29851 + }, + { + "epoch": 23.843450479233226, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 29852 + }, + { + "epoch": 23.844249201277954, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 29853 + }, + { + "epoch": 23.845047923322685, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 29854 + }, + { + "epoch": 23.845846645367413, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 29855 + }, + { + "epoch": 23.84664536741214, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 29856 + }, + { + "epoch": 23.847444089456868, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 29857 + }, + { + "epoch": 23.8482428115016, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 29858 + }, + { + "epoch": 23.849041533546327, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 29859 + }, + { + "epoch": 23.849840255591054, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 29860 + }, + { + "epoch": 23.850638977635782, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 29861 + }, + { + "epoch": 23.85143769968051, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 29862 + }, + { + "epoch": 23.85223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 29863 + }, + { + "epoch": 23.85303514376997, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 29864 + }, + { + "epoch": 23.853833865814696, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 29865 + }, + { + "epoch": 23.854632587859424, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 29866 + }, + { + "epoch": 23.855431309904155, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 29867 + }, + { + "epoch": 23.856230031948883, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 29868 + }, + { + "epoch": 23.85702875399361, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29869 + }, + { + "epoch": 23.857827476038338, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 29870 + }, + { + "epoch": 23.858626198083066, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 29871 + }, + { + "epoch": 23.859424920127797, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 29872 + }, + { + "epoch": 23.860223642172524, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 29873 + }, + { + "epoch": 23.861022364217252, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 29874 + }, + { + "epoch": 23.86182108626198, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29875 + }, + { + "epoch": 23.86261980830671, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 29876 + }, + { + "epoch": 23.86341853035144, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29877 + }, + { + "epoch": 23.864217252396166, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 29878 + }, + { + "epoch": 23.865015974440894, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 29879 + }, + { + "epoch": 23.86581469648562, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 29880 + }, + { + "epoch": 23.866613418530353, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 29881 + }, + { + "epoch": 23.86741214057508, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 29882 + }, + { + "epoch": 23.868210862619808, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29883 + }, + { + "epoch": 23.869009584664536, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 29884 + }, + { + "epoch": 23.869808306709267, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 29885 + }, + { + "epoch": 23.870607028753994, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 29886 + }, + { + "epoch": 23.871405750798722, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 29887 + }, + { + "epoch": 23.87220447284345, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 29888 + }, + { + "epoch": 23.873003194888177, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 29889 + }, + { + "epoch": 23.87380191693291, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 29890 + }, + { + "epoch": 23.874600638977636, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 29891 + }, + { + "epoch": 23.875399361022364, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 29892 + }, + { + "epoch": 23.87619808306709, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 29893 + }, + { + "epoch": 23.876996805111823, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 29894 + }, + { + "epoch": 23.87779552715655, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 29895 + }, + { + "epoch": 23.878594249201278, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 29896 + }, + { + "epoch": 23.879392971246006, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 29897 + }, + { + "epoch": 23.880191693290733, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29898 + }, + { + "epoch": 23.880990415335464, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 29899 + }, + { + "epoch": 23.881789137380192, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 29900 + }, + { + "epoch": 23.88258785942492, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 29901 + }, + { + "epoch": 23.883386581469647, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 29902 + }, + { + "epoch": 23.88418530351438, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 29903 + }, + { + "epoch": 23.884984025559106, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 29904 + }, + { + "epoch": 23.885782747603834, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 29905 + }, + { + "epoch": 23.88658146964856, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 29906 + }, + { + "epoch": 23.88738019169329, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 29907 + }, + { + "epoch": 23.88817891373802, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 29908 + }, + { + "epoch": 23.888977635782748, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 29909 + }, + { + "epoch": 23.889776357827476, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 29910 + }, + { + "epoch": 23.890575079872203, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 29911 + }, + { + "epoch": 23.891373801916934, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 29912 + }, + { + "epoch": 23.892172523961662, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 29913 + }, + { + "epoch": 23.89297124600639, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 29914 + }, + { + "epoch": 23.893769968051117, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 29915 + }, + { + "epoch": 23.894568690095845, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 29916 + }, + { + "epoch": 23.895367412140576, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 29917 + }, + { + "epoch": 23.896166134185304, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0714, + "step": 29918 + }, + { + "epoch": 23.89696485623003, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29919 + }, + { + "epoch": 23.89776357827476, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 29920 + }, + { + "epoch": 23.89856230031949, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 29921 + }, + { + "epoch": 23.899361022364218, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 29922 + }, + { + "epoch": 23.900159744408946, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 29923 + }, + { + "epoch": 23.900958466453673, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 29924 + }, + { + "epoch": 23.9017571884984, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 29925 + }, + { + "epoch": 23.902555910543132, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 29926 + }, + { + "epoch": 23.90335463258786, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 29927 + }, + { + "epoch": 23.904153354632587, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 29928 + }, + { + "epoch": 23.904952076677315, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 29929 + }, + { + "epoch": 23.905750798722046, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 29930 + }, + { + "epoch": 23.906549520766774, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 29931 + }, + { + "epoch": 23.9073482428115, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 29932 + }, + { + "epoch": 23.90814696485623, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 29933 + }, + { + "epoch": 23.908945686900957, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 29934 + }, + { + "epoch": 23.909744408945688, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 29935 + }, + { + "epoch": 23.910543130990416, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 29936 + }, + { + "epoch": 23.911341853035143, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 29937 + }, + { + "epoch": 23.91214057507987, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 29938 + }, + { + "epoch": 23.912939297124602, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 29939 + }, + { + "epoch": 23.91373801916933, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 29940 + }, + { + "epoch": 23.914536741214057, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 29941 + }, + { + "epoch": 23.915335463258785, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 29942 + }, + { + "epoch": 23.916134185303516, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 29943 + }, + { + "epoch": 23.916932907348244, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 29944 + }, + { + "epoch": 23.91773162939297, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 29945 + }, + { + "epoch": 23.9185303514377, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 29946 + }, + { + "epoch": 23.919329073482427, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 29947 + }, + { + "epoch": 23.920127795527158, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 29948 + }, + { + "epoch": 23.920926517571885, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 29949 + }, + { + "epoch": 23.921725239616613, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 29950 + }, + { + "epoch": 23.92252396166134, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 29951 + }, + { + "epoch": 23.923322683706072, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 29952 + }, + { + "epoch": 23.9241214057508, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 29953 + }, + { + "epoch": 23.924920127795527, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 29954 + }, + { + "epoch": 23.925718849840255, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 29955 + }, + { + "epoch": 23.926517571884983, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 29956 + }, + { + "epoch": 23.927316293929714, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 29957 + }, + { + "epoch": 23.92811501597444, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 29958 + }, + { + "epoch": 23.92891373801917, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 29959 + }, + { + "epoch": 23.929712460063897, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 29960 + }, + { + "epoch": 23.930511182108628, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 29961 + }, + { + "epoch": 23.931309904153355, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 29962 + }, + { + "epoch": 23.932108626198083, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 29963 + }, + { + "epoch": 23.93290734824281, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 29964 + }, + { + "epoch": 23.93370607028754, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 29965 + }, + { + "epoch": 23.93450479233227, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 29966 + }, + { + "epoch": 23.935303514376997, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 29967 + }, + { + "epoch": 23.936102236421725, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 29968 + }, + { + "epoch": 23.936900958466452, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 29969 + }, + { + "epoch": 23.937699680511184, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 29970 + }, + { + "epoch": 23.93849840255591, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 29971 + }, + { + "epoch": 23.93929712460064, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 29972 + }, + { + "epoch": 23.940095846645367, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 29973 + }, + { + "epoch": 23.940894568690094, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 29974 + }, + { + "epoch": 23.941693290734825, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 29975 + }, + { + "epoch": 23.942492012779553, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 29976 + }, + { + "epoch": 23.94329073482428, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 29977 + }, + { + "epoch": 23.94408945686901, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 29978 + }, + { + "epoch": 23.94488817891374, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 29979 + }, + { + "epoch": 23.945686900958467, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 29980 + }, + { + "epoch": 23.946485623003195, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 29981 + }, + { + "epoch": 23.947284345047922, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0728, + "step": 29982 + }, + { + "epoch": 23.94808306709265, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 29983 + }, + { + "epoch": 23.94888178913738, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 29984 + }, + { + "epoch": 23.94968051118211, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 29985 + }, + { + "epoch": 23.950479233226837, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 29986 + }, + { + "epoch": 23.951277955271564, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 29987 + }, + { + "epoch": 23.952076677316295, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 29988 + }, + { + "epoch": 23.952875399361023, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 29989 + }, + { + "epoch": 23.95367412140575, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 29990 + }, + { + "epoch": 23.95447284345048, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 29991 + }, + { + "epoch": 23.955271565495206, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 29992 + }, + { + "epoch": 23.956070287539937, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 29993 + }, + { + "epoch": 23.956869009584665, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 29994 + }, + { + "epoch": 23.957667731629392, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 29995 + }, + { + "epoch": 23.95846645367412, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 29996 + }, + { + "epoch": 23.95926517571885, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 29997 + }, + { + "epoch": 23.96006389776358, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 29998 + }, + { + "epoch": 23.960862619808307, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 29999 + }, + { + "epoch": 23.961661341853034, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 30000 + }, + { + "epoch": 23.962460063897765, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 30001 + }, + { + "epoch": 23.963258785942493, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 30002 + }, + { + "epoch": 23.96405750798722, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 30003 + }, + { + "epoch": 23.96485623003195, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30004 + }, + { + "epoch": 23.965654952076676, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30005 + }, + { + "epoch": 23.966453674121407, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 30006 + }, + { + "epoch": 23.967252396166135, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 30007 + }, + { + "epoch": 23.968051118210862, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 30008 + }, + { + "epoch": 23.96884984025559, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 30009 + }, + { + "epoch": 23.96964856230032, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 30010 + }, + { + "epoch": 23.97044728434505, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 30011 + }, + { + "epoch": 23.971246006389777, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30012 + }, + { + "epoch": 23.972044728434504, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 30013 + }, + { + "epoch": 23.972843450479232, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 30014 + }, + { + "epoch": 23.973642172523963, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 30015 + }, + { + "epoch": 23.97444089456869, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 30016 + }, + { + "epoch": 23.97523961661342, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 30017 + }, + { + "epoch": 23.976038338658146, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 30018 + }, + { + "epoch": 23.976837060702877, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 30019 + }, + { + "epoch": 23.977635782747605, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 30020 + }, + { + "epoch": 23.978434504792332, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 30021 + }, + { + "epoch": 23.97923322683706, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30022 + }, + { + "epoch": 23.980031948881788, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 30023 + }, + { + "epoch": 23.98083067092652, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30024 + }, + { + "epoch": 23.981629392971247, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 30025 + }, + { + "epoch": 23.982428115015974, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 30026 + }, + { + "epoch": 23.983226837060702, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 30027 + }, + { + "epoch": 23.984025559105433, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 30028 + }, + { + "epoch": 23.98482428115016, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 30029 + }, + { + "epoch": 23.98562300319489, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 30030 + }, + { + "epoch": 23.986421725239616, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 30031 + }, + { + "epoch": 23.987220447284344, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 30032 + }, + { + "epoch": 23.988019169329075, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 30033 + }, + { + "epoch": 23.988817891373802, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 30034 + }, + { + "epoch": 23.98961661341853, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 30035 + }, + { + "epoch": 23.990415335463258, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 30036 + }, + { + "epoch": 23.99121405750799, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 30037 + }, + { + "epoch": 23.992012779552716, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 30038 + }, + { + "epoch": 23.992811501597444, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 30039 + }, + { + "epoch": 23.99361022364217, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30040 + }, + { + "epoch": 23.9944089456869, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 30041 + }, + { + "epoch": 23.99520766773163, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30042 + }, + { + "epoch": 23.99600638977636, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 30043 + }, + { + "epoch": 23.996805111821086, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 30044 + }, + { + "epoch": 23.997603833865814, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 30045 + }, + { + "epoch": 23.998402555910545, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 30046 + }, + { + "epoch": 23.999201277955272, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 30047 + }, + { + "epoch": 24.0, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 30048 + }, + { + "epoch": 24.000798722044728, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 30049 + }, + { + "epoch": 24.001597444089455, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 30050 + }, + { + "epoch": 24.002396166134186, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30051 + }, + { + "epoch": 24.003194888178914, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 30052 + }, + { + "epoch": 24.00399361022364, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 30053 + }, + { + "epoch": 24.00479233226837, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 30054 + }, + { + "epoch": 24.0055910543131, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 30055 + }, + { + "epoch": 24.00638977635783, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 30056 + }, + { + "epoch": 24.007188498402556, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 30057 + }, + { + "epoch": 24.007987220447284, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30058 + }, + { + "epoch": 24.00878594249201, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 30059 + }, + { + "epoch": 24.009584664536742, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 30060 + }, + { + "epoch": 24.01038338658147, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 30061 + }, + { + "epoch": 24.011182108626198, + "grad_norm": 0.328125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 30062 + }, + { + "epoch": 24.011980830670925, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 30063 + }, + { + "epoch": 24.012779552715656, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 30064 + }, + { + "epoch": 24.013578274760384, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 30065 + }, + { + "epoch": 24.01437699680511, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 30066 + }, + { + "epoch": 24.01517571884984, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 30067 + }, + { + "epoch": 24.015974440894567, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 30068 + }, + { + "epoch": 24.016773162939298, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 30069 + }, + { + "epoch": 24.017571884984026, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 30070 + }, + { + "epoch": 24.018370607028753, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30071 + }, + { + "epoch": 24.01916932907348, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1018, + "step": 30072 + }, + { + "epoch": 24.019968051118212, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 30073 + }, + { + "epoch": 24.02076677316294, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 30074 + }, + { + "epoch": 24.021565495207668, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 30075 + }, + { + "epoch": 24.022364217252395, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 30076 + }, + { + "epoch": 24.023162939297123, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 30077 + }, + { + "epoch": 24.023961661341854, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 30078 + }, + { + "epoch": 24.02476038338658, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 30079 + }, + { + "epoch": 24.02555910543131, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 30080 + }, + { + "epoch": 24.026357827476037, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 30081 + }, + { + "epoch": 24.027156549520768, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 30082 + }, + { + "epoch": 24.027955271565496, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30083 + }, + { + "epoch": 24.028753993610223, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 30084 + }, + { + "epoch": 24.02955271565495, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 30085 + }, + { + "epoch": 24.03035143769968, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 30086 + }, + { + "epoch": 24.03115015974441, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 30087 + }, + { + "epoch": 24.031948881789138, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 30088 + }, + { + "epoch": 24.032747603833865, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30089 + }, + { + "epoch": 24.033546325878593, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30090 + }, + { + "epoch": 24.034345047923324, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 30091 + }, + { + "epoch": 24.03514376996805, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30092 + }, + { + "epoch": 24.03594249201278, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0738, + "step": 30093 + }, + { + "epoch": 24.036741214057507, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 30094 + }, + { + "epoch": 24.037539936102238, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 30095 + }, + { + "epoch": 24.038338658146966, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 30096 + }, + { + "epoch": 24.039137380191693, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 30097 + }, + { + "epoch": 24.03993610223642, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30098 + }, + { + "epoch": 24.04073482428115, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30099 + }, + { + "epoch": 24.04153354632588, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 30100 + }, + { + "epoch": 24.042332268370608, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 30101 + }, + { + "epoch": 24.043130990415335, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 30102 + }, + { + "epoch": 24.043929712460063, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 30103 + }, + { + "epoch": 24.044728434504794, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30104 + }, + { + "epoch": 24.04552715654952, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 30105 + }, + { + "epoch": 24.04632587859425, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 30106 + }, + { + "epoch": 24.047124600638977, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 30107 + }, + { + "epoch": 24.047923322683705, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 30108 + }, + { + "epoch": 24.048722044728436, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 30109 + }, + { + "epoch": 24.049520766773163, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 30110 + }, + { + "epoch": 24.05031948881789, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 30111 + }, + { + "epoch": 24.05111821086262, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 30112 + }, + { + "epoch": 24.05191693290735, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 30113 + }, + { + "epoch": 24.052715654952078, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 30114 + }, + { + "epoch": 24.053514376996805, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 30115 + }, + { + "epoch": 24.054313099041533, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 30116 + }, + { + "epoch": 24.05511182108626, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 30117 + }, + { + "epoch": 24.05591054313099, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 30118 + }, + { + "epoch": 24.05670926517572, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 30119 + }, + { + "epoch": 24.057507987220447, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 30120 + }, + { + "epoch": 24.058306709265175, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 30121 + }, + { + "epoch": 24.059105431309906, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30122 + }, + { + "epoch": 24.059904153354633, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 30123 + }, + { + "epoch": 24.06070287539936, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 30124 + }, + { + "epoch": 24.06150159744409, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 30125 + }, + { + "epoch": 24.062300319488816, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 30126 + }, + { + "epoch": 24.063099041533548, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 30127 + }, + { + "epoch": 24.063897763578275, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 30128 + }, + { + "epoch": 24.064696485623003, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 30129 + }, + { + "epoch": 24.06549520766773, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 30130 + }, + { + "epoch": 24.06629392971246, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0773, + "step": 30131 + }, + { + "epoch": 24.06709265175719, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 30132 + }, + { + "epoch": 24.067891373801917, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 30133 + }, + { + "epoch": 24.068690095846645, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30134 + }, + { + "epoch": 24.069488817891372, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 30135 + }, + { + "epoch": 24.070287539936103, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 30136 + }, + { + "epoch": 24.07108626198083, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 30137 + }, + { + "epoch": 24.07188498402556, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 30138 + }, + { + "epoch": 24.072683706070286, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 30139 + }, + { + "epoch": 24.073482428115017, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 30140 + }, + { + "epoch": 24.074281150159745, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30141 + }, + { + "epoch": 24.075079872204473, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 30142 + }, + { + "epoch": 24.0758785942492, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 30143 + }, + { + "epoch": 24.076677316293928, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 30144 + }, + { + "epoch": 24.07747603833866, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 30145 + }, + { + "epoch": 24.078274760383387, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0773, + "step": 30146 + }, + { + "epoch": 24.079073482428115, + "grad_norm": 0.349609375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 30147 + }, + { + "epoch": 24.079872204472842, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 30148 + }, + { + "epoch": 24.080670926517573, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 30149 + }, + { + "epoch": 24.0814696485623, + "grad_norm": 0.31640625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30150 + }, + { + "epoch": 24.08226837060703, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 30151 + }, + { + "epoch": 24.083067092651756, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 30152 + }, + { + "epoch": 24.083865814696484, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 30153 + }, + { + "epoch": 24.084664536741215, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 30154 + }, + { + "epoch": 24.085463258785943, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 30155 + }, + { + "epoch": 24.08626198083067, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 30156 + }, + { + "epoch": 24.087060702875398, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 30157 + }, + { + "epoch": 24.08785942492013, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 30158 + }, + { + "epoch": 24.088658146964857, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 30159 + }, + { + "epoch": 24.089456869009584, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 30160 + }, + { + "epoch": 24.090255591054312, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 30161 + }, + { + "epoch": 24.09105431309904, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 30162 + }, + { + "epoch": 24.09185303514377, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 30163 + }, + { + "epoch": 24.0926517571885, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 30164 + }, + { + "epoch": 24.093450479233226, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 30165 + }, + { + "epoch": 24.094249201277954, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 30166 + }, + { + "epoch": 24.095047923322685, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 30167 + }, + { + "epoch": 24.095846645367413, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 30168 + }, + { + "epoch": 24.09664536741214, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 30169 + }, + { + "epoch": 24.097444089456868, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 30170 + }, + { + "epoch": 24.0982428115016, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 30171 + }, + { + "epoch": 24.099041533546327, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 30172 + }, + { + "epoch": 24.099840255591054, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 30173 + }, + { + "epoch": 24.100638977635782, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 30174 + }, + { + "epoch": 24.10143769968051, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30175 + }, + { + "epoch": 24.10223642172524, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 30176 + }, + { + "epoch": 24.10303514376997, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 30177 + }, + { + "epoch": 24.103833865814696, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 30178 + }, + { + "epoch": 24.104632587859424, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 30179 + }, + { + "epoch": 24.105431309904155, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 30180 + }, + { + "epoch": 24.106230031948883, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 30181 + }, + { + "epoch": 24.10702875399361, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 30182 + }, + { + "epoch": 24.107827476038338, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0743, + "step": 30183 + }, + { + "epoch": 24.108626198083066, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 30184 + }, + { + "epoch": 24.109424920127797, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 30185 + }, + { + "epoch": 24.110223642172524, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30186 + }, + { + "epoch": 24.111022364217252, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30187 + }, + { + "epoch": 24.11182108626198, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0773, + "step": 30188 + }, + { + "epoch": 24.11261980830671, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 30189 + }, + { + "epoch": 24.11341853035144, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 30190 + }, + { + "epoch": 24.114217252396166, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 30191 + }, + { + "epoch": 24.115015974440894, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 30192 + }, + { + "epoch": 24.11581469648562, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 30193 + }, + { + "epoch": 24.116613418530353, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 30194 + }, + { + "epoch": 24.11741214057508, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 30195 + }, + { + "epoch": 24.118210862619808, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 30196 + }, + { + "epoch": 24.119009584664536, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 30197 + }, + { + "epoch": 24.119808306709267, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 30198 + }, + { + "epoch": 24.120607028753994, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 30199 + }, + { + "epoch": 24.121405750798722, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 30200 + }, + { + "epoch": 24.12220447284345, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 30201 + }, + { + "epoch": 24.123003194888177, + "grad_norm": 0.09619140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 30202 + }, + { + "epoch": 24.12380191693291, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 30203 + }, + { + "epoch": 24.124600638977636, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 30204 + }, + { + "epoch": 24.125399361022364, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 30205 + }, + { + "epoch": 24.12619808306709, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 30206 + }, + { + "epoch": 24.126996805111823, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 30207 + }, + { + "epoch": 24.12779552715655, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 30208 + }, + { + "epoch": 24.128594249201278, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30209 + }, + { + "epoch": 24.129392971246006, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 30210 + }, + { + "epoch": 24.130191693290733, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 30211 + }, + { + "epoch": 24.130990415335464, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 30212 + }, + { + "epoch": 24.131789137380192, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 30213 + }, + { + "epoch": 24.13258785942492, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 30214 + }, + { + "epoch": 24.133386581469647, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 30215 + }, + { + "epoch": 24.13418530351438, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 30216 + }, + { + "epoch": 24.134984025559106, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 30217 + }, + { + "epoch": 24.135782747603834, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 30218 + }, + { + "epoch": 24.13658146964856, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30219 + }, + { + "epoch": 24.13738019169329, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 30220 + }, + { + "epoch": 24.13817891373802, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 30221 + }, + { + "epoch": 24.138977635782748, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 30222 + }, + { + "epoch": 24.139776357827476, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 30223 + }, + { + "epoch": 24.140575079872203, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 30224 + }, + { + "epoch": 24.141373801916934, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 30225 + }, + { + "epoch": 24.142172523961662, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 30226 + }, + { + "epoch": 24.14297124600639, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 30227 + }, + { + "epoch": 24.143769968051117, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 30228 + }, + { + "epoch": 24.144568690095845, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 30229 + }, + { + "epoch": 24.145367412140576, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30230 + }, + { + "epoch": 24.146166134185304, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 30231 + }, + { + "epoch": 24.14696485623003, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 30232 + }, + { + "epoch": 24.14776357827476, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 30233 + }, + { + "epoch": 24.14856230031949, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 30234 + }, + { + "epoch": 24.149361022364218, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 30235 + }, + { + "epoch": 24.150159744408946, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 30236 + }, + { + "epoch": 24.150958466453673, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 30237 + }, + { + "epoch": 24.1517571884984, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 30238 + }, + { + "epoch": 24.152555910543132, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 30239 + }, + { + "epoch": 24.15335463258786, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 30240 + }, + { + "epoch": 24.154153354632587, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 30241 + }, + { + "epoch": 24.154952076677315, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30242 + }, + { + "epoch": 24.155750798722046, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 30243 + }, + { + "epoch": 24.156549520766774, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 30244 + }, + { + "epoch": 24.1573482428115, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 30245 + }, + { + "epoch": 24.15814696485623, + "grad_norm": 0.09814453125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30246 + }, + { + "epoch": 24.15894568690096, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 30247 + }, + { + "epoch": 24.159744408945688, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 30248 + }, + { + "epoch": 24.160543130990416, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 30249 + }, + { + "epoch": 24.161341853035143, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 30250 + }, + { + "epoch": 24.16214057507987, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 30251 + }, + { + "epoch": 24.162939297124602, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 30252 + }, + { + "epoch": 24.16373801916933, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 30253 + }, + { + "epoch": 24.164536741214057, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 30254 + }, + { + "epoch": 24.165335463258785, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 30255 + }, + { + "epoch": 24.166134185303516, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 30256 + }, + { + "epoch": 24.166932907348244, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 30257 + }, + { + "epoch": 24.16773162939297, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 30258 + }, + { + "epoch": 24.1685303514377, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 30259 + }, + { + "epoch": 24.169329073482427, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 30260 + }, + { + "epoch": 24.170127795527158, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30261 + }, + { + "epoch": 24.170926517571885, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 30262 + }, + { + "epoch": 24.171725239616613, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 30263 + }, + { + "epoch": 24.17252396166134, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 30264 + }, + { + "epoch": 24.173322683706072, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 30265 + }, + { + "epoch": 24.1741214057508, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 30266 + }, + { + "epoch": 24.174920127795527, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 30267 + }, + { + "epoch": 24.175718849840255, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 30268 + }, + { + "epoch": 24.176517571884983, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 30269 + }, + { + "epoch": 24.177316293929714, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 30270 + }, + { + "epoch": 24.17811501597444, + "grad_norm": 0.326171875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30271 + }, + { + "epoch": 24.17891373801917, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 30272 + }, + { + "epoch": 24.179712460063897, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 30273 + }, + { + "epoch": 24.180511182108628, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 30274 + }, + { + "epoch": 24.181309904153355, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 30275 + }, + { + "epoch": 24.182108626198083, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 30276 + }, + { + "epoch": 24.18290734824281, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 30277 + }, + { + "epoch": 24.18370607028754, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 30278 + }, + { + "epoch": 24.18450479233227, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 30279 + }, + { + "epoch": 24.185303514376997, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 30280 + }, + { + "epoch": 24.186102236421725, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30281 + }, + { + "epoch": 24.186900958466452, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 30282 + }, + { + "epoch": 24.187699680511184, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 30283 + }, + { + "epoch": 24.18849840255591, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30284 + }, + { + "epoch": 24.18929712460064, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 30285 + }, + { + "epoch": 24.190095846645367, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 30286 + }, + { + "epoch": 24.190894568690094, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 30287 + }, + { + "epoch": 24.191693290734825, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 30288 + }, + { + "epoch": 24.192492012779553, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 30289 + }, + { + "epoch": 24.19329073482428, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30290 + }, + { + "epoch": 24.19408945686901, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 30291 + }, + { + "epoch": 24.19488817891374, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 30292 + }, + { + "epoch": 24.195686900958467, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 30293 + }, + { + "epoch": 24.196485623003195, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30294 + }, + { + "epoch": 24.197284345047922, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 30295 + }, + { + "epoch": 24.19808306709265, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 30296 + }, + { + "epoch": 24.19888178913738, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 30297 + }, + { + "epoch": 24.19968051118211, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 30298 + }, + { + "epoch": 24.200479233226837, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30299 + }, + { + "epoch": 24.201277955271564, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 30300 + }, + { + "epoch": 24.202076677316295, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 30301 + }, + { + "epoch": 24.202875399361023, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 30302 + }, + { + "epoch": 24.20367412140575, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 30303 + }, + { + "epoch": 24.20447284345048, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 30304 + }, + { + "epoch": 24.205271565495206, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30305 + }, + { + "epoch": 24.206070287539937, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 30306 + }, + { + "epoch": 24.206869009584665, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 30307 + }, + { + "epoch": 24.207667731629392, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 30308 + }, + { + "epoch": 24.20846645367412, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 30309 + }, + { + "epoch": 24.20926517571885, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 30310 + }, + { + "epoch": 24.21006389776358, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 30311 + }, + { + "epoch": 24.210862619808307, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 30312 + }, + { + "epoch": 24.211661341853034, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 30313 + }, + { + "epoch": 24.212460063897762, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 30314 + }, + { + "epoch": 24.213258785942493, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 30315 + }, + { + "epoch": 24.21405750798722, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30316 + }, + { + "epoch": 24.21485623003195, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30317 + }, + { + "epoch": 24.215654952076676, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 30318 + }, + { + "epoch": 24.216453674121407, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1037, + "step": 30319 + }, + { + "epoch": 24.217252396166135, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 30320 + }, + { + "epoch": 24.218051118210862, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 30321 + }, + { + "epoch": 24.21884984025559, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30322 + }, + { + "epoch": 24.21964856230032, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 30323 + }, + { + "epoch": 24.22044728434505, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 30324 + }, + { + "epoch": 24.221246006389777, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30325 + }, + { + "epoch": 24.222044728434504, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 30326 + }, + { + "epoch": 24.222843450479232, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 30327 + }, + { + "epoch": 24.223642172523963, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 30328 + }, + { + "epoch": 24.22444089456869, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30329 + }, + { + "epoch": 24.22523961661342, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 30330 + }, + { + "epoch": 24.226038338658146, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 30331 + }, + { + "epoch": 24.226837060702877, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 30332 + }, + { + "epoch": 24.227635782747605, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30333 + }, + { + "epoch": 24.228434504792332, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 30334 + }, + { + "epoch": 24.22923322683706, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 30335 + }, + { + "epoch": 24.230031948881788, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 30336 + }, + { + "epoch": 24.23083067092652, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 30337 + }, + { + "epoch": 24.231629392971247, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 30338 + }, + { + "epoch": 24.232428115015974, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 30339 + }, + { + "epoch": 24.233226837060702, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 30340 + }, + { + "epoch": 24.234025559105433, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 30341 + }, + { + "epoch": 24.23482428115016, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 30342 + }, + { + "epoch": 24.23562300319489, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 30343 + }, + { + "epoch": 24.236421725239616, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 30344 + }, + { + "epoch": 24.237220447284344, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 30345 + }, + { + "epoch": 24.238019169329075, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 30346 + }, + { + "epoch": 24.238817891373802, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 30347 + }, + { + "epoch": 24.23961661341853, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30348 + }, + { + "epoch": 24.240415335463258, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 30349 + }, + { + "epoch": 24.24121405750799, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 30350 + }, + { + "epoch": 24.242012779552716, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 30351 + }, + { + "epoch": 24.242811501597444, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 30352 + }, + { + "epoch": 24.24361022364217, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 30353 + }, + { + "epoch": 24.2444089456869, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 30354 + }, + { + "epoch": 24.24520766773163, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 30355 + }, + { + "epoch": 24.24600638977636, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 30356 + }, + { + "epoch": 24.246805111821086, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30357 + }, + { + "epoch": 24.247603833865814, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30358 + }, + { + "epoch": 24.248402555910545, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 30359 + }, + { + "epoch": 24.249201277955272, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 30360 + }, + { + "epoch": 24.25, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 30361 + }, + { + "epoch": 24.250798722044728, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30362 + }, + { + "epoch": 24.251597444089455, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 30363 + }, + { + "epoch": 24.252396166134186, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 30364 + }, + { + "epoch": 24.253194888178914, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 30365 + }, + { + "epoch": 24.25399361022364, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 30366 + }, + { + "epoch": 24.25479233226837, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 30367 + }, + { + "epoch": 24.2555910543131, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 30368 + }, + { + "epoch": 24.25638977635783, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 30369 + }, + { + "epoch": 24.257188498402556, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 30370 + }, + { + "epoch": 24.257987220447284, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 30371 + }, + { + "epoch": 24.25878594249201, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 30372 + }, + { + "epoch": 24.259584664536742, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 30373 + }, + { + "epoch": 24.26038338658147, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 30374 + }, + { + "epoch": 24.261182108626198, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 30375 + }, + { + "epoch": 24.261980830670925, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 30376 + }, + { + "epoch": 24.262779552715656, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 30377 + }, + { + "epoch": 24.263578274760384, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 30378 + }, + { + "epoch": 24.26437699680511, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 30379 + }, + { + "epoch": 24.26517571884984, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 30380 + }, + { + "epoch": 24.265974440894567, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30381 + }, + { + "epoch": 24.266773162939298, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 30382 + }, + { + "epoch": 24.267571884984026, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 30383 + }, + { + "epoch": 24.268370607028753, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 30384 + }, + { + "epoch": 24.26916932907348, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30385 + }, + { + "epoch": 24.269968051118212, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30386 + }, + { + "epoch": 24.27076677316294, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 30387 + }, + { + "epoch": 24.271565495207668, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 30388 + }, + { + "epoch": 24.272364217252395, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 30389 + }, + { + "epoch": 24.273162939297123, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 30390 + }, + { + "epoch": 24.273961661341854, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 30391 + }, + { + "epoch": 24.27476038338658, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 30392 + }, + { + "epoch": 24.27555910543131, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 30393 + }, + { + "epoch": 24.276357827476037, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 30394 + }, + { + "epoch": 24.277156549520768, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 30395 + }, + { + "epoch": 24.277955271565496, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 30396 + }, + { + "epoch": 24.278753993610223, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 30397 + }, + { + "epoch": 24.27955271565495, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 30398 + }, + { + "epoch": 24.28035143769968, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 30399 + }, + { + "epoch": 24.28115015974441, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 30400 + }, + { + "epoch": 24.281948881789138, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 30401 + }, + { + "epoch": 24.282747603833865, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 30402 + }, + { + "epoch": 24.283546325878593, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 30403 + }, + { + "epoch": 24.284345047923324, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30404 + }, + { + "epoch": 24.28514376996805, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30405 + }, + { + "epoch": 24.28594249201278, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 30406 + }, + { + "epoch": 24.286741214057507, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 30407 + }, + { + "epoch": 24.287539936102238, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 30408 + }, + { + "epoch": 24.288338658146966, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 30409 + }, + { + "epoch": 24.289137380191693, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 30410 + }, + { + "epoch": 24.28993610223642, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 30411 + }, + { + "epoch": 24.29073482428115, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 30412 + }, + { + "epoch": 24.29153354632588, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 30413 + }, + { + "epoch": 24.292332268370608, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 30414 + }, + { + "epoch": 24.293130990415335, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 30415 + }, + { + "epoch": 24.293929712460063, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 30416 + }, + { + "epoch": 24.294728434504794, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 30417 + }, + { + "epoch": 24.29552715654952, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 30418 + }, + { + "epoch": 24.29632587859425, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 30419 + }, + { + "epoch": 24.297124600638977, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 30420 + }, + { + "epoch": 24.297923322683705, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 30421 + }, + { + "epoch": 24.298722044728436, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 30422 + }, + { + "epoch": 24.299520766773163, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 30423 + }, + { + "epoch": 24.30031948881789, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 30424 + }, + { + "epoch": 24.30111821086262, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 30425 + }, + { + "epoch": 24.30191693290735, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30426 + }, + { + "epoch": 24.302715654952078, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 30427 + }, + { + "epoch": 24.303514376996805, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 30428 + }, + { + "epoch": 24.304313099041533, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 30429 + }, + { + "epoch": 24.30511182108626, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 30430 + }, + { + "epoch": 24.30591054313099, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30431 + }, + { + "epoch": 24.30670926517572, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 30432 + }, + { + "epoch": 24.307507987220447, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 30433 + }, + { + "epoch": 24.308306709265175, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30434 + }, + { + "epoch": 24.309105431309906, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 30435 + }, + { + "epoch": 24.309904153354633, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 30436 + }, + { + "epoch": 24.31070287539936, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30437 + }, + { + "epoch": 24.31150159744409, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30438 + }, + { + "epoch": 24.312300319488816, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 30439 + }, + { + "epoch": 24.313099041533548, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 30440 + }, + { + "epoch": 24.313897763578275, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 30441 + }, + { + "epoch": 24.314696485623003, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 30442 + }, + { + "epoch": 24.31549520766773, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 30443 + }, + { + "epoch": 24.31629392971246, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 30444 + }, + { + "epoch": 24.31709265175719, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 30445 + }, + { + "epoch": 24.317891373801917, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 30446 + }, + { + "epoch": 24.318690095846645, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30447 + }, + { + "epoch": 24.319488817891372, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 30448 + }, + { + "epoch": 24.320287539936103, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 30449 + }, + { + "epoch": 24.32108626198083, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 30450 + }, + { + "epoch": 24.32188498402556, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 30451 + }, + { + "epoch": 24.322683706070286, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0748, + "step": 30452 + }, + { + "epoch": 24.323482428115017, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30453 + }, + { + "epoch": 24.324281150159745, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 30454 + }, + { + "epoch": 24.325079872204473, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 30455 + }, + { + "epoch": 24.3258785942492, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 30456 + }, + { + "epoch": 24.326677316293928, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 30457 + }, + { + "epoch": 24.32747603833866, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 30458 + }, + { + "epoch": 24.328274760383387, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 30459 + }, + { + "epoch": 24.329073482428115, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 30460 + }, + { + "epoch": 24.329872204472842, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 30461 + }, + { + "epoch": 24.330670926517573, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 30462 + }, + { + "epoch": 24.3314696485623, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 30463 + }, + { + "epoch": 24.33226837060703, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 30464 + }, + { + "epoch": 24.333067092651756, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 30465 + }, + { + "epoch": 24.333865814696484, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30466 + }, + { + "epoch": 24.334664536741215, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 30467 + }, + { + "epoch": 24.335463258785943, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 30468 + }, + { + "epoch": 24.33626198083067, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 30469 + }, + { + "epoch": 24.337060702875398, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 30470 + }, + { + "epoch": 24.33785942492013, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30471 + }, + { + "epoch": 24.338658146964857, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 30472 + }, + { + "epoch": 24.339456869009584, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 30473 + }, + { + "epoch": 24.340255591054312, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 30474 + }, + { + "epoch": 24.341054313099043, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 30475 + }, + { + "epoch": 24.34185303514377, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 30476 + }, + { + "epoch": 24.3426517571885, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30477 + }, + { + "epoch": 24.343450479233226, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 30478 + }, + { + "epoch": 24.344249201277954, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 30479 + }, + { + "epoch": 24.345047923322685, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 30480 + }, + { + "epoch": 24.345846645367413, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 30481 + }, + { + "epoch": 24.34664536741214, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 30482 + }, + { + "epoch": 24.347444089456868, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 30483 + }, + { + "epoch": 24.3482428115016, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 30484 + }, + { + "epoch": 24.349041533546327, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 30485 + }, + { + "epoch": 24.349840255591054, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 30486 + }, + { + "epoch": 24.350638977635782, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 30487 + }, + { + "epoch": 24.35143769968051, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 30488 + }, + { + "epoch": 24.35223642172524, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1076, + "step": 30489 + }, + { + "epoch": 24.35303514376997, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 30490 + }, + { + "epoch": 24.353833865814696, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 30491 + }, + { + "epoch": 24.354632587859424, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 30492 + }, + { + "epoch": 24.355431309904155, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 30493 + }, + { + "epoch": 24.356230031948883, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 30494 + }, + { + "epoch": 24.35702875399361, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 30495 + }, + { + "epoch": 24.357827476038338, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 30496 + }, + { + "epoch": 24.358626198083066, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30497 + }, + { + "epoch": 24.359424920127797, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 30498 + }, + { + "epoch": 24.360223642172524, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 30499 + }, + { + "epoch": 24.361022364217252, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 30500 + }, + { + "epoch": 24.36182108626198, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 30501 + }, + { + "epoch": 24.36261980830671, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 30502 + }, + { + "epoch": 24.36341853035144, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30503 + }, + { + "epoch": 24.364217252396166, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 30504 + }, + { + "epoch": 24.365015974440894, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 30505 + }, + { + "epoch": 24.36581469648562, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 30506 + }, + { + "epoch": 24.366613418530353, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 30507 + }, + { + "epoch": 24.36741214057508, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 30508 + }, + { + "epoch": 24.368210862619808, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30509 + }, + { + "epoch": 24.369009584664536, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30510 + }, + { + "epoch": 24.369808306709267, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 30511 + }, + { + "epoch": 24.370607028753994, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30512 + }, + { + "epoch": 24.371405750798722, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 30513 + }, + { + "epoch": 24.37220447284345, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 30514 + }, + { + "epoch": 24.373003194888177, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 30515 + }, + { + "epoch": 24.37380191693291, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 30516 + }, + { + "epoch": 24.374600638977636, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 30517 + }, + { + "epoch": 24.375399361022364, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 30518 + }, + { + "epoch": 24.37619808306709, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 30519 + }, + { + "epoch": 24.376996805111823, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 30520 + }, + { + "epoch": 24.37779552715655, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 30521 + }, + { + "epoch": 24.378594249201278, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 30522 + }, + { + "epoch": 24.379392971246006, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 30523 + }, + { + "epoch": 24.380191693290733, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 30524 + }, + { + "epoch": 24.380990415335464, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 30525 + }, + { + "epoch": 24.381789137380192, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 30526 + }, + { + "epoch": 24.38258785942492, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30527 + }, + { + "epoch": 24.383386581469647, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 30528 + }, + { + "epoch": 24.38418530351438, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 30529 + }, + { + "epoch": 24.384984025559106, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 30530 + }, + { + "epoch": 24.385782747603834, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30531 + }, + { + "epoch": 24.38658146964856, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30532 + }, + { + "epoch": 24.38738019169329, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 30533 + }, + { + "epoch": 24.38817891373802, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30534 + }, + { + "epoch": 24.388977635782748, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 30535 + }, + { + "epoch": 24.389776357827476, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 30536 + }, + { + "epoch": 24.390575079872203, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 30537 + }, + { + "epoch": 24.391373801916934, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 30538 + }, + { + "epoch": 24.392172523961662, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 30539 + }, + { + "epoch": 24.39297124600639, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 30540 + }, + { + "epoch": 24.393769968051117, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 30541 + }, + { + "epoch": 24.394568690095845, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 30542 + }, + { + "epoch": 24.395367412140576, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 30543 + }, + { + "epoch": 24.396166134185304, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 30544 + }, + { + "epoch": 24.39696485623003, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30545 + }, + { + "epoch": 24.39776357827476, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 30546 + }, + { + "epoch": 24.39856230031949, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30547 + }, + { + "epoch": 24.399361022364218, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 30548 + }, + { + "epoch": 24.400159744408946, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 30549 + }, + { + "epoch": 24.400958466453673, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 30550 + }, + { + "epoch": 24.4017571884984, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30551 + }, + { + "epoch": 24.402555910543132, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 30552 + }, + { + "epoch": 24.40335463258786, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 30553 + }, + { + "epoch": 24.404153354632587, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 30554 + }, + { + "epoch": 24.404952076677315, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30555 + }, + { + "epoch": 24.405750798722046, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 30556 + }, + { + "epoch": 24.406549520766774, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30557 + }, + { + "epoch": 24.4073482428115, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 30558 + }, + { + "epoch": 24.40814696485623, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 30559 + }, + { + "epoch": 24.408945686900957, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 30560 + }, + { + "epoch": 24.409744408945688, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 30561 + }, + { + "epoch": 24.410543130990416, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 30562 + }, + { + "epoch": 24.411341853035143, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 30563 + }, + { + "epoch": 24.41214057507987, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 30564 + }, + { + "epoch": 24.412939297124602, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 30565 + }, + { + "epoch": 24.41373801916933, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 30566 + }, + { + "epoch": 24.414536741214057, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 30567 + }, + { + "epoch": 24.415335463258785, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 30568 + }, + { + "epoch": 24.416134185303516, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30569 + }, + { + "epoch": 24.416932907348244, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 30570 + }, + { + "epoch": 24.41773162939297, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 30571 + }, + { + "epoch": 24.4185303514377, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30572 + }, + { + "epoch": 24.419329073482427, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 30573 + }, + { + "epoch": 24.420127795527158, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 30574 + }, + { + "epoch": 24.420926517571885, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 30575 + }, + { + "epoch": 24.421725239616613, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 30576 + }, + { + "epoch": 24.42252396166134, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30577 + }, + { + "epoch": 24.423322683706072, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 30578 + }, + { + "epoch": 24.4241214057508, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30579 + }, + { + "epoch": 24.424920127795527, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 30580 + }, + { + "epoch": 24.425718849840255, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 30581 + }, + { + "epoch": 24.426517571884983, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 30582 + }, + { + "epoch": 24.427316293929714, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 30583 + }, + { + "epoch": 24.42811501597444, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30584 + }, + { + "epoch": 24.42891373801917, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 30585 + }, + { + "epoch": 24.429712460063897, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 30586 + }, + { + "epoch": 24.430511182108628, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 30587 + }, + { + "epoch": 24.431309904153355, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 30588 + }, + { + "epoch": 24.432108626198083, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 30589 + }, + { + "epoch": 24.43290734824281, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 30590 + }, + { + "epoch": 24.43370607028754, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 30591 + }, + { + "epoch": 24.43450479233227, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30592 + }, + { + "epoch": 24.435303514376997, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 30593 + }, + { + "epoch": 24.436102236421725, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30594 + }, + { + "epoch": 24.436900958466452, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 30595 + }, + { + "epoch": 24.437699680511184, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 30596 + }, + { + "epoch": 24.43849840255591, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 30597 + }, + { + "epoch": 24.43929712460064, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 30598 + }, + { + "epoch": 24.440095846645367, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30599 + }, + { + "epoch": 24.440894568690094, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 30600 + }, + { + "epoch": 24.441693290734825, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 30601 + }, + { + "epoch": 24.442492012779553, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 30602 + }, + { + "epoch": 24.44329073482428, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 30603 + }, + { + "epoch": 24.44408945686901, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 30604 + }, + { + "epoch": 24.44488817891374, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1048, + "step": 30605 + }, + { + "epoch": 24.445686900958467, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30606 + }, + { + "epoch": 24.446485623003195, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 30607 + }, + { + "epoch": 24.447284345047922, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 30608 + }, + { + "epoch": 24.44808306709265, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 30609 + }, + { + "epoch": 24.44888178913738, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 30610 + }, + { + "epoch": 24.44968051118211, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 30611 + }, + { + "epoch": 24.450479233226837, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 30612 + }, + { + "epoch": 24.451277955271564, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 30613 + }, + { + "epoch": 24.452076677316295, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 30614 + }, + { + "epoch": 24.452875399361023, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 30615 + }, + { + "epoch": 24.45367412140575, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 30616 + }, + { + "epoch": 24.45447284345048, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 30617 + }, + { + "epoch": 24.455271565495206, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 30618 + }, + { + "epoch": 24.456070287539937, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 30619 + }, + { + "epoch": 24.456869009584665, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 30620 + }, + { + "epoch": 24.457667731629392, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 30621 + }, + { + "epoch": 24.45846645367412, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 30622 + }, + { + "epoch": 24.45926517571885, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 30623 + }, + { + "epoch": 24.46006389776358, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 30624 + }, + { + "epoch": 24.460862619808307, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 30625 + }, + { + "epoch": 24.461661341853034, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 30626 + }, + { + "epoch": 24.462460063897762, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30627 + }, + { + "epoch": 24.463258785942493, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 30628 + }, + { + "epoch": 24.46405750798722, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 30629 + }, + { + "epoch": 24.46485623003195, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 30630 + }, + { + "epoch": 24.465654952076676, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 30631 + }, + { + "epoch": 24.466453674121407, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 30632 + }, + { + "epoch": 24.467252396166135, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30633 + }, + { + "epoch": 24.468051118210862, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 30634 + }, + { + "epoch": 24.46884984025559, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 30635 + }, + { + "epoch": 24.46964856230032, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 30636 + }, + { + "epoch": 24.47044728434505, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 30637 + }, + { + "epoch": 24.471246006389777, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 30638 + }, + { + "epoch": 24.472044728434504, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 30639 + }, + { + "epoch": 24.472843450479232, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 30640 + }, + { + "epoch": 24.473642172523963, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 30641 + }, + { + "epoch": 24.47444089456869, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 30642 + }, + { + "epoch": 24.47523961661342, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 30643 + }, + { + "epoch": 24.476038338658146, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 30644 + }, + { + "epoch": 24.476837060702877, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 30645 + }, + { + "epoch": 24.477635782747605, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30646 + }, + { + "epoch": 24.478434504792332, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 30647 + }, + { + "epoch": 24.47923322683706, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 30648 + }, + { + "epoch": 24.480031948881788, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 30649 + }, + { + "epoch": 24.48083067092652, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 30650 + }, + { + "epoch": 24.481629392971247, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 30651 + }, + { + "epoch": 24.482428115015974, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 30652 + }, + { + "epoch": 24.483226837060702, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 30653 + }, + { + "epoch": 24.484025559105433, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 30654 + }, + { + "epoch": 24.48482428115016, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 30655 + }, + { + "epoch": 24.48562300319489, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 30656 + }, + { + "epoch": 24.486421725239616, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 30657 + }, + { + "epoch": 24.487220447284344, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 30658 + }, + { + "epoch": 24.488019169329075, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 30659 + }, + { + "epoch": 24.488817891373802, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 30660 + }, + { + "epoch": 24.48961661341853, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 30661 + }, + { + "epoch": 24.490415335463258, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30662 + }, + { + "epoch": 24.49121405750799, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 30663 + }, + { + "epoch": 24.492012779552716, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 30664 + }, + { + "epoch": 24.492811501597444, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30665 + }, + { + "epoch": 24.49361022364217, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 30666 + }, + { + "epoch": 24.4944089456869, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30667 + }, + { + "epoch": 24.49520766773163, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 30668 + }, + { + "epoch": 24.49600638977636, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 30669 + }, + { + "epoch": 24.496805111821086, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 30670 + }, + { + "epoch": 24.497603833865814, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 30671 + }, + { + "epoch": 24.498402555910545, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 30672 + }, + { + "epoch": 24.499201277955272, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 30673 + }, + { + "epoch": 24.5, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30674 + }, + { + "epoch": 24.500798722044728, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 30675 + }, + { + "epoch": 24.501597444089455, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 30676 + }, + { + "epoch": 24.502396166134186, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 30677 + }, + { + "epoch": 24.503194888178914, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 30678 + }, + { + "epoch": 24.50399361022364, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 30679 + }, + { + "epoch": 24.50479233226837, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 30680 + }, + { + "epoch": 24.5055910543131, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 30681 + }, + { + "epoch": 24.50638977635783, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 30682 + }, + { + "epoch": 24.507188498402556, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 30683 + }, + { + "epoch": 24.507987220447284, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 30684 + }, + { + "epoch": 24.50878594249201, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 30685 + }, + { + "epoch": 24.509584664536742, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 30686 + }, + { + "epoch": 24.51038338658147, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 30687 + }, + { + "epoch": 24.511182108626198, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 30688 + }, + { + "epoch": 24.511980830670925, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0743, + "step": 30689 + }, + { + "epoch": 24.512779552715656, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 30690 + }, + { + "epoch": 24.513578274760384, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 30691 + }, + { + "epoch": 24.51437699680511, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30692 + }, + { + "epoch": 24.51517571884984, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 30693 + }, + { + "epoch": 24.515974440894567, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 30694 + }, + { + "epoch": 24.516773162939298, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 30695 + }, + { + "epoch": 24.517571884984026, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 30696 + }, + { + "epoch": 24.518370607028753, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 30697 + }, + { + "epoch": 24.51916932907348, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 30698 + }, + { + "epoch": 24.519968051118212, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 30699 + }, + { + "epoch": 24.52076677316294, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 30700 + }, + { + "epoch": 24.521565495207668, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 30701 + }, + { + "epoch": 24.522364217252395, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 30702 + }, + { + "epoch": 24.523162939297123, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30703 + }, + { + "epoch": 24.523961661341854, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 30704 + }, + { + "epoch": 24.52476038338658, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30705 + }, + { + "epoch": 24.52555910543131, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 30706 + }, + { + "epoch": 24.526357827476037, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 30707 + }, + { + "epoch": 24.527156549520768, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 30708 + }, + { + "epoch": 24.527955271565496, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 30709 + }, + { + "epoch": 24.528753993610223, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 30710 + }, + { + "epoch": 24.52955271565495, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30711 + }, + { + "epoch": 24.53035143769968, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 30712 + }, + { + "epoch": 24.53115015974441, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30713 + }, + { + "epoch": 24.531948881789138, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 30714 + }, + { + "epoch": 24.532747603833865, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 30715 + }, + { + "epoch": 24.533546325878593, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 30716 + }, + { + "epoch": 24.534345047923324, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30717 + }, + { + "epoch": 24.53514376996805, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 30718 + }, + { + "epoch": 24.53594249201278, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 30719 + }, + { + "epoch": 24.536741214057507, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 30720 + }, + { + "epoch": 24.537539936102235, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 30721 + }, + { + "epoch": 24.538338658146966, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 30722 + }, + { + "epoch": 24.539137380191693, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 30723 + }, + { + "epoch": 24.53993610223642, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 30724 + }, + { + "epoch": 24.54073482428115, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 30725 + }, + { + "epoch": 24.54153354632588, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30726 + }, + { + "epoch": 24.542332268370608, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 30727 + }, + { + "epoch": 24.543130990415335, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 30728 + }, + { + "epoch": 24.543929712460063, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30729 + }, + { + "epoch": 24.544728434504794, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 30730 + }, + { + "epoch": 24.54552715654952, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 30731 + }, + { + "epoch": 24.54632587859425, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 30732 + }, + { + "epoch": 24.547124600638977, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30733 + }, + { + "epoch": 24.547923322683705, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 30734 + }, + { + "epoch": 24.548722044728436, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30735 + }, + { + "epoch": 24.549520766773163, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 30736 + }, + { + "epoch": 24.55031948881789, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30737 + }, + { + "epoch": 24.55111821086262, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 30738 + }, + { + "epoch": 24.55191693290735, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 30739 + }, + { + "epoch": 24.552715654952078, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 30740 + }, + { + "epoch": 24.553514376996805, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 30741 + }, + { + "epoch": 24.554313099041533, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 30742 + }, + { + "epoch": 24.55511182108626, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30743 + }, + { + "epoch": 24.55591054313099, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 30744 + }, + { + "epoch": 24.55670926517572, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 30745 + }, + { + "epoch": 24.557507987220447, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 30746 + }, + { + "epoch": 24.558306709265175, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 30747 + }, + { + "epoch": 24.559105431309906, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 30748 + }, + { + "epoch": 24.559904153354633, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 30749 + }, + { + "epoch": 24.56070287539936, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 30750 + }, + { + "epoch": 24.56150159744409, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 30751 + }, + { + "epoch": 24.562300319488816, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 30752 + }, + { + "epoch": 24.563099041533548, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 30753 + }, + { + "epoch": 24.563897763578275, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 30754 + }, + { + "epoch": 24.564696485623003, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 30755 + }, + { + "epoch": 24.56549520766773, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 30756 + }, + { + "epoch": 24.56629392971246, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 30757 + }, + { + "epoch": 24.56709265175719, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30758 + }, + { + "epoch": 24.567891373801917, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 30759 + }, + { + "epoch": 24.568690095846645, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 30760 + }, + { + "epoch": 24.569488817891372, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 30761 + }, + { + "epoch": 24.570287539936103, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 30762 + }, + { + "epoch": 24.57108626198083, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 30763 + }, + { + "epoch": 24.57188498402556, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30764 + }, + { + "epoch": 24.572683706070286, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30765 + }, + { + "epoch": 24.573482428115017, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30766 + }, + { + "epoch": 24.574281150159745, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 30767 + }, + { + "epoch": 24.575079872204473, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 30768 + }, + { + "epoch": 24.5758785942492, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 30769 + }, + { + "epoch": 24.576677316293928, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 30770 + }, + { + "epoch": 24.57747603833866, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 30771 + }, + { + "epoch": 24.578274760383387, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 30772 + }, + { + "epoch": 24.579073482428115, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 30773 + }, + { + "epoch": 24.579872204472842, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 30774 + }, + { + "epoch": 24.580670926517573, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 30775 + }, + { + "epoch": 24.5814696485623, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 30776 + }, + { + "epoch": 24.58226837060703, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 30777 + }, + { + "epoch": 24.583067092651756, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30778 + }, + { + "epoch": 24.583865814696484, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 30779 + }, + { + "epoch": 24.584664536741215, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 30780 + }, + { + "epoch": 24.585463258785943, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 30781 + }, + { + "epoch": 24.58626198083067, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 30782 + }, + { + "epoch": 24.587060702875398, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 30783 + }, + { + "epoch": 24.58785942492013, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 30784 + }, + { + "epoch": 24.588658146964857, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 30785 + }, + { + "epoch": 24.589456869009584, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 30786 + }, + { + "epoch": 24.590255591054312, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 30787 + }, + { + "epoch": 24.591054313099043, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 30788 + }, + { + "epoch": 24.59185303514377, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 30789 + }, + { + "epoch": 24.5926517571885, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 30790 + }, + { + "epoch": 24.593450479233226, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 30791 + }, + { + "epoch": 24.594249201277954, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 30792 + }, + { + "epoch": 24.595047923322685, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 30793 + }, + { + "epoch": 24.595846645367413, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30794 + }, + { + "epoch": 24.59664536741214, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 30795 + }, + { + "epoch": 24.597444089456868, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 30796 + }, + { + "epoch": 24.5982428115016, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 30797 + }, + { + "epoch": 24.599041533546327, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 30798 + }, + { + "epoch": 24.599840255591054, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 30799 + }, + { + "epoch": 24.600638977635782, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 30800 + }, + { + "epoch": 24.60143769968051, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 30801 + }, + { + "epoch": 24.60223642172524, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 30802 + }, + { + "epoch": 24.60303514376997, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 30803 + }, + { + "epoch": 24.603833865814696, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 30804 + }, + { + "epoch": 24.604632587859424, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 30805 + }, + { + "epoch": 24.605431309904155, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 30806 + }, + { + "epoch": 24.606230031948883, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30807 + }, + { + "epoch": 24.60702875399361, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 30808 + }, + { + "epoch": 24.607827476038338, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 30809 + }, + { + "epoch": 24.608626198083066, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 30810 + }, + { + "epoch": 24.609424920127797, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 30811 + }, + { + "epoch": 24.610223642172524, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30812 + }, + { + "epoch": 24.611022364217252, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 30813 + }, + { + "epoch": 24.61182108626198, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 30814 + }, + { + "epoch": 24.61261980830671, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 30815 + }, + { + "epoch": 24.61341853035144, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 30816 + }, + { + "epoch": 24.614217252396166, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 30817 + }, + { + "epoch": 24.615015974440894, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 30818 + }, + { + "epoch": 24.61581469648562, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 30819 + }, + { + "epoch": 24.616613418530353, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 30820 + }, + { + "epoch": 24.61741214057508, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 30821 + }, + { + "epoch": 24.618210862619808, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 30822 + }, + { + "epoch": 24.619009584664536, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 30823 + }, + { + "epoch": 24.619808306709267, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 30824 + }, + { + "epoch": 24.620607028753994, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.1013, + "step": 30825 + }, + { + "epoch": 24.621405750798722, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 30826 + }, + { + "epoch": 24.62220447284345, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 30827 + }, + { + "epoch": 24.623003194888177, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 30828 + }, + { + "epoch": 24.62380191693291, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 30829 + }, + { + "epoch": 24.624600638977636, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 30830 + }, + { + "epoch": 24.625399361022364, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 30831 + }, + { + "epoch": 24.62619808306709, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 30832 + }, + { + "epoch": 24.626996805111823, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 30833 + }, + { + "epoch": 24.62779552715655, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 30834 + }, + { + "epoch": 24.628594249201278, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 30835 + }, + { + "epoch": 24.629392971246006, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 30836 + }, + { + "epoch": 24.630191693290733, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 30837 + }, + { + "epoch": 24.630990415335464, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 30838 + }, + { + "epoch": 24.631789137380192, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 30839 + }, + { + "epoch": 24.63258785942492, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 30840 + }, + { + "epoch": 24.633386581469647, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 30841 + }, + { + "epoch": 24.63418530351438, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 30842 + }, + { + "epoch": 24.634984025559106, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 30843 + }, + { + "epoch": 24.635782747603834, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 30844 + }, + { + "epoch": 24.63658146964856, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 30845 + }, + { + "epoch": 24.63738019169329, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 30846 + }, + { + "epoch": 24.63817891373802, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 30847 + }, + { + "epoch": 24.638977635782748, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30848 + }, + { + "epoch": 24.639776357827476, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.1069, + "step": 30849 + }, + { + "epoch": 24.640575079872203, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 30850 + }, + { + "epoch": 24.641373801916934, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 30851 + }, + { + "epoch": 24.642172523961662, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 30852 + }, + { + "epoch": 24.64297124600639, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 30853 + }, + { + "epoch": 24.643769968051117, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 30854 + }, + { + "epoch": 24.644568690095845, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 30855 + }, + { + "epoch": 24.645367412140576, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 30856 + }, + { + "epoch": 24.646166134185304, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 30857 + }, + { + "epoch": 24.64696485623003, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 30858 + }, + { + "epoch": 24.64776357827476, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 30859 + }, + { + "epoch": 24.64856230031949, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 30860 + }, + { + "epoch": 24.649361022364218, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 30861 + }, + { + "epoch": 24.650159744408946, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 30862 + }, + { + "epoch": 24.650958466453673, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 30863 + }, + { + "epoch": 24.6517571884984, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 30864 + }, + { + "epoch": 24.652555910543132, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 30865 + }, + { + "epoch": 24.65335463258786, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 30866 + }, + { + "epoch": 24.654153354632587, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 30867 + }, + { + "epoch": 24.654952076677315, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 30868 + }, + { + "epoch": 24.655750798722046, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30869 + }, + { + "epoch": 24.656549520766774, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 30870 + }, + { + "epoch": 24.6573482428115, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 30871 + }, + { + "epoch": 24.65814696485623, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 30872 + }, + { + "epoch": 24.658945686900957, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 30873 + }, + { + "epoch": 24.659744408945688, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30874 + }, + { + "epoch": 24.660543130990416, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30875 + }, + { + "epoch": 24.661341853035143, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 30876 + }, + { + "epoch": 24.66214057507987, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 30877 + }, + { + "epoch": 24.662939297124602, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 30878 + }, + { + "epoch": 24.66373801916933, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 30879 + }, + { + "epoch": 24.664536741214057, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 30880 + }, + { + "epoch": 24.665335463258785, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 30881 + }, + { + "epoch": 24.666134185303516, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 30882 + }, + { + "epoch": 24.666932907348244, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 30883 + }, + { + "epoch": 24.66773162939297, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 30884 + }, + { + "epoch": 24.6685303514377, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 30885 + }, + { + "epoch": 24.669329073482427, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 30886 + }, + { + "epoch": 24.670127795527158, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 30887 + }, + { + "epoch": 24.670926517571885, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 30888 + }, + { + "epoch": 24.671725239616613, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 30889 + }, + { + "epoch": 24.67252396166134, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 30890 + }, + { + "epoch": 24.673322683706072, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 30891 + }, + { + "epoch": 24.6741214057508, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 30892 + }, + { + "epoch": 24.674920127795527, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 30893 + }, + { + "epoch": 24.675718849840255, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 30894 + }, + { + "epoch": 24.676517571884983, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30895 + }, + { + "epoch": 24.677316293929714, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30896 + }, + { + "epoch": 24.67811501597444, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 30897 + }, + { + "epoch": 24.67891373801917, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 30898 + }, + { + "epoch": 24.679712460063897, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30899 + }, + { + "epoch": 24.680511182108628, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 30900 + }, + { + "epoch": 24.681309904153355, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 30901 + }, + { + "epoch": 24.682108626198083, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 30902 + }, + { + "epoch": 24.68290734824281, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 30903 + }, + { + "epoch": 24.68370607028754, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 30904 + }, + { + "epoch": 24.68450479233227, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 30905 + }, + { + "epoch": 24.685303514376997, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 30906 + }, + { + "epoch": 24.686102236421725, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 30907 + }, + { + "epoch": 24.686900958466452, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 30908 + }, + { + "epoch": 24.687699680511184, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 30909 + }, + { + "epoch": 24.68849840255591, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 30910 + }, + { + "epoch": 24.68929712460064, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30911 + }, + { + "epoch": 24.690095846645367, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 30912 + }, + { + "epoch": 24.690894568690094, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 30913 + }, + { + "epoch": 24.691693290734825, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 30914 + }, + { + "epoch": 24.692492012779553, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 30915 + }, + { + "epoch": 24.69329073482428, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 30916 + }, + { + "epoch": 24.69408945686901, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 30917 + }, + { + "epoch": 24.69488817891374, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30918 + }, + { + "epoch": 24.695686900958467, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 30919 + }, + { + "epoch": 24.696485623003195, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 30920 + }, + { + "epoch": 24.697284345047922, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 30921 + }, + { + "epoch": 24.69808306709265, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 30922 + }, + { + "epoch": 24.69888178913738, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 30923 + }, + { + "epoch": 24.69968051118211, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30924 + }, + { + "epoch": 24.700479233226837, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 30925 + }, + { + "epoch": 24.701277955271564, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 30926 + }, + { + "epoch": 24.702076677316295, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 30927 + }, + { + "epoch": 24.702875399361023, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 30928 + }, + { + "epoch": 24.70367412140575, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 30929 + }, + { + "epoch": 24.70447284345048, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 30930 + }, + { + "epoch": 24.705271565495206, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 30931 + }, + { + "epoch": 24.706070287539937, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 30932 + }, + { + "epoch": 24.706869009584665, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 30933 + }, + { + "epoch": 24.707667731629392, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 30934 + }, + { + "epoch": 24.70846645367412, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 30935 + }, + { + "epoch": 24.70926517571885, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 30936 + }, + { + "epoch": 24.71006389776358, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 30937 + }, + { + "epoch": 24.710862619808307, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 30938 + }, + { + "epoch": 24.711661341853034, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 30939 + }, + { + "epoch": 24.712460063897765, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 30940 + }, + { + "epoch": 24.713258785942493, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 30941 + }, + { + "epoch": 24.71405750798722, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 30942 + }, + { + "epoch": 24.71485623003195, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 30943 + }, + { + "epoch": 24.715654952076676, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30944 + }, + { + "epoch": 24.716453674121407, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 30945 + }, + { + "epoch": 24.717252396166135, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 30946 + }, + { + "epoch": 24.718051118210862, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30947 + }, + { + "epoch": 24.71884984025559, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 30948 + }, + { + "epoch": 24.71964856230032, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 30949 + }, + { + "epoch": 24.72044728434505, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 30950 + }, + { + "epoch": 24.721246006389777, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 30951 + }, + { + "epoch": 24.722044728434504, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 30952 + }, + { + "epoch": 24.722843450479232, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 30953 + }, + { + "epoch": 24.723642172523963, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 30954 + }, + { + "epoch": 24.72444089456869, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 30955 + }, + { + "epoch": 24.72523961661342, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 30956 + }, + { + "epoch": 24.726038338658146, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30957 + }, + { + "epoch": 24.726837060702877, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 30958 + }, + { + "epoch": 24.727635782747605, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30959 + }, + { + "epoch": 24.728434504792332, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 30960 + }, + { + "epoch": 24.72923322683706, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 30961 + }, + { + "epoch": 24.730031948881788, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 30962 + }, + { + "epoch": 24.73083067092652, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 30963 + }, + { + "epoch": 24.731629392971247, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 30964 + }, + { + "epoch": 24.732428115015974, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 30965 + }, + { + "epoch": 24.733226837060702, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 30966 + }, + { + "epoch": 24.734025559105433, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 30967 + }, + { + "epoch": 24.73482428115016, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 30968 + }, + { + "epoch": 24.73562300319489, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 30969 + }, + { + "epoch": 24.736421725239616, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 30970 + }, + { + "epoch": 24.737220447284344, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 30971 + }, + { + "epoch": 24.738019169329075, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 30972 + }, + { + "epoch": 24.738817891373802, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 30973 + }, + { + "epoch": 24.73961661341853, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 30974 + }, + { + "epoch": 24.740415335463258, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 30975 + }, + { + "epoch": 24.74121405750799, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 30976 + }, + { + "epoch": 24.742012779552716, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30977 + }, + { + "epoch": 24.742811501597444, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 30978 + }, + { + "epoch": 24.74361022364217, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 30979 + }, + { + "epoch": 24.7444089456869, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 30980 + }, + { + "epoch": 24.74520766773163, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 30981 + }, + { + "epoch": 24.74600638977636, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 30982 + }, + { + "epoch": 24.746805111821086, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 30983 + }, + { + "epoch": 24.747603833865814, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 30984 + }, + { + "epoch": 24.748402555910545, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 30985 + }, + { + "epoch": 24.749201277955272, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 30986 + }, + { + "epoch": 24.75, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 30987 + }, + { + "epoch": 24.750798722044728, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 30988 + }, + { + "epoch": 24.751597444089455, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 30989 + }, + { + "epoch": 24.752396166134186, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 30990 + }, + { + "epoch": 24.753194888178914, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 30991 + }, + { + "epoch": 24.75399361022364, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 30992 + }, + { + "epoch": 24.75479233226837, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 30993 + }, + { + "epoch": 24.7555910543131, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 30994 + }, + { + "epoch": 24.75638977635783, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 30995 + }, + { + "epoch": 24.757188498402556, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 30996 + }, + { + "epoch": 24.757987220447284, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 30997 + }, + { + "epoch": 24.75878594249201, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 30998 + }, + { + "epoch": 24.759584664536742, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 30999 + }, + { + "epoch": 24.76038338658147, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 31000 + }, + { + "epoch": 24.761182108626198, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 31001 + }, + { + "epoch": 24.761980830670925, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0753, + "step": 31002 + }, + { + "epoch": 24.762779552715656, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 31003 + }, + { + "epoch": 24.763578274760384, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 31004 + }, + { + "epoch": 24.76437699680511, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 31005 + }, + { + "epoch": 24.76517571884984, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0723, + "step": 31006 + }, + { + "epoch": 24.765974440894567, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 31007 + }, + { + "epoch": 24.766773162939298, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 31008 + }, + { + "epoch": 24.767571884984026, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 31009 + }, + { + "epoch": 24.768370607028753, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 31010 + }, + { + "epoch": 24.76916932907348, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 31011 + }, + { + "epoch": 24.769968051118212, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 31012 + }, + { + "epoch": 24.77076677316294, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 31013 + }, + { + "epoch": 24.771565495207668, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 31014 + }, + { + "epoch": 24.772364217252395, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 31015 + }, + { + "epoch": 24.773162939297123, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 31016 + }, + { + "epoch": 24.773961661341854, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 31017 + }, + { + "epoch": 24.77476038338658, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 31018 + }, + { + "epoch": 24.77555910543131, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 31019 + }, + { + "epoch": 24.776357827476037, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 31020 + }, + { + "epoch": 24.777156549520768, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 31021 + }, + { + "epoch": 24.777955271565496, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 31022 + }, + { + "epoch": 24.778753993610223, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31023 + }, + { + "epoch": 24.77955271565495, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 31024 + }, + { + "epoch": 24.78035143769968, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 31025 + }, + { + "epoch": 24.78115015974441, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 31026 + }, + { + "epoch": 24.781948881789138, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 31027 + }, + { + "epoch": 24.782747603833865, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 31028 + }, + { + "epoch": 24.783546325878593, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 31029 + }, + { + "epoch": 24.784345047923324, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 31030 + }, + { + "epoch": 24.78514376996805, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 31031 + }, + { + "epoch": 24.78594249201278, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 31032 + }, + { + "epoch": 24.786741214057507, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31033 + }, + { + "epoch": 24.787539936102235, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 31034 + }, + { + "epoch": 24.788338658146966, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 31035 + }, + { + "epoch": 24.789137380191693, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 31036 + }, + { + "epoch": 24.78993610223642, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 31037 + }, + { + "epoch": 24.79073482428115, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 31038 + }, + { + "epoch": 24.79153354632588, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 31039 + }, + { + "epoch": 24.792332268370608, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 31040 + }, + { + "epoch": 24.793130990415335, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 31041 + }, + { + "epoch": 24.793929712460063, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 31042 + }, + { + "epoch": 24.794728434504794, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 31043 + }, + { + "epoch": 24.79552715654952, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31044 + }, + { + "epoch": 24.79632587859425, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 31045 + }, + { + "epoch": 24.797124600638977, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 31046 + }, + { + "epoch": 24.797923322683705, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 31047 + }, + { + "epoch": 24.798722044728436, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 31048 + }, + { + "epoch": 24.799520766773163, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 31049 + }, + { + "epoch": 24.80031948881789, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 31050 + }, + { + "epoch": 24.80111821086262, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 31051 + }, + { + "epoch": 24.80191693290735, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 31052 + }, + { + "epoch": 24.802715654952078, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 31053 + }, + { + "epoch": 24.803514376996805, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 31054 + }, + { + "epoch": 24.804313099041533, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 31055 + }, + { + "epoch": 24.80511182108626, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 31056 + }, + { + "epoch": 24.80591054313099, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 31057 + }, + { + "epoch": 24.80670926517572, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31058 + }, + { + "epoch": 24.807507987220447, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 31059 + }, + { + "epoch": 24.808306709265175, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 31060 + }, + { + "epoch": 24.809105431309906, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 31061 + }, + { + "epoch": 24.809904153354633, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 31062 + }, + { + "epoch": 24.81070287539936, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 31063 + }, + { + "epoch": 24.81150159744409, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 31064 + }, + { + "epoch": 24.812300319488816, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 31065 + }, + { + "epoch": 24.813099041533548, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 31066 + }, + { + "epoch": 24.813897763578275, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31067 + }, + { + "epoch": 24.814696485623003, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31068 + }, + { + "epoch": 24.81549520766773, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 31069 + }, + { + "epoch": 24.81629392971246, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 31070 + }, + { + "epoch": 24.81709265175719, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 31071 + }, + { + "epoch": 24.817891373801917, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 31072 + }, + { + "epoch": 24.818690095846645, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31073 + }, + { + "epoch": 24.819488817891372, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31074 + }, + { + "epoch": 24.820287539936103, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 31075 + }, + { + "epoch": 24.82108626198083, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 31076 + }, + { + "epoch": 24.82188498402556, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 31077 + }, + { + "epoch": 24.822683706070286, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 31078 + }, + { + "epoch": 24.823482428115017, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 31079 + }, + { + "epoch": 24.824281150159745, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 31080 + }, + { + "epoch": 24.825079872204473, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 31081 + }, + { + "epoch": 24.8258785942492, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 31082 + }, + { + "epoch": 24.826677316293928, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 31083 + }, + { + "epoch": 24.82747603833866, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 31084 + }, + { + "epoch": 24.828274760383387, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 31085 + }, + { + "epoch": 24.829073482428115, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 31086 + }, + { + "epoch": 24.829872204472842, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 31087 + }, + { + "epoch": 24.830670926517573, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 31088 + }, + { + "epoch": 24.8314696485623, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 31089 + }, + { + "epoch": 24.83226837060703, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 31090 + }, + { + "epoch": 24.833067092651756, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 31091 + }, + { + "epoch": 24.833865814696484, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 31092 + }, + { + "epoch": 24.834664536741215, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 31093 + }, + { + "epoch": 24.835463258785943, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 31094 + }, + { + "epoch": 24.83626198083067, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 31095 + }, + { + "epoch": 24.837060702875398, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 31096 + }, + { + "epoch": 24.83785942492013, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 31097 + }, + { + "epoch": 24.838658146964857, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 31098 + }, + { + "epoch": 24.839456869009584, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 31099 + }, + { + "epoch": 24.840255591054312, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 31100 + }, + { + "epoch": 24.841054313099043, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 31101 + }, + { + "epoch": 24.84185303514377, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 31102 + }, + { + "epoch": 24.8426517571885, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 31103 + }, + { + "epoch": 24.843450479233226, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 31104 + }, + { + "epoch": 24.844249201277954, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 31105 + }, + { + "epoch": 24.845047923322685, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 31106 + }, + { + "epoch": 24.845846645367413, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 31107 + }, + { + "epoch": 24.84664536741214, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 31108 + }, + { + "epoch": 24.847444089456868, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 31109 + }, + { + "epoch": 24.8482428115016, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 31110 + }, + { + "epoch": 24.849041533546327, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 31111 + }, + { + "epoch": 24.849840255591054, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 31112 + }, + { + "epoch": 24.850638977635782, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 31113 + }, + { + "epoch": 24.85143769968051, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 31114 + }, + { + "epoch": 24.85223642172524, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 31115 + }, + { + "epoch": 24.85303514376997, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 31116 + }, + { + "epoch": 24.853833865814696, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 31117 + }, + { + "epoch": 24.854632587859424, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 31118 + }, + { + "epoch": 24.855431309904155, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 31119 + }, + { + "epoch": 24.856230031948883, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 31120 + }, + { + "epoch": 24.85702875399361, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 31121 + }, + { + "epoch": 24.857827476038338, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 31122 + }, + { + "epoch": 24.858626198083066, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 31123 + }, + { + "epoch": 24.859424920127797, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 31124 + }, + { + "epoch": 24.860223642172524, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 31125 + }, + { + "epoch": 24.861022364217252, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31126 + }, + { + "epoch": 24.86182108626198, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 31127 + }, + { + "epoch": 24.86261980830671, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 31128 + }, + { + "epoch": 24.86341853035144, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 31129 + }, + { + "epoch": 24.864217252396166, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31130 + }, + { + "epoch": 24.865015974440894, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 31131 + }, + { + "epoch": 24.86581469648562, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 31132 + }, + { + "epoch": 24.866613418530353, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 31133 + }, + { + "epoch": 24.86741214057508, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 31134 + }, + { + "epoch": 24.868210862619808, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 31135 + }, + { + "epoch": 24.869009584664536, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 31136 + }, + { + "epoch": 24.869808306709267, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 31137 + }, + { + "epoch": 24.870607028753994, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 31138 + }, + { + "epoch": 24.871405750798722, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 31139 + }, + { + "epoch": 24.87220447284345, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 31140 + }, + { + "epoch": 24.873003194888177, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 31141 + }, + { + "epoch": 24.87380191693291, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 31142 + }, + { + "epoch": 24.874600638977636, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 31143 + }, + { + "epoch": 24.875399361022364, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0765, + "step": 31144 + }, + { + "epoch": 24.87619808306709, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 31145 + }, + { + "epoch": 24.876996805111823, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 31146 + }, + { + "epoch": 24.87779552715655, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 31147 + }, + { + "epoch": 24.878594249201278, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 31148 + }, + { + "epoch": 24.879392971246006, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31149 + }, + { + "epoch": 24.880191693290733, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 31150 + }, + { + "epoch": 24.880990415335464, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 31151 + }, + { + "epoch": 24.881789137380192, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 31152 + }, + { + "epoch": 24.88258785942492, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0743, + "step": 31153 + }, + { + "epoch": 24.883386581469647, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 31154 + }, + { + "epoch": 24.88418530351438, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 31155 + }, + { + "epoch": 24.884984025559106, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 31156 + }, + { + "epoch": 24.885782747603834, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 31157 + }, + { + "epoch": 24.88658146964856, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 31158 + }, + { + "epoch": 24.88738019169329, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 31159 + }, + { + "epoch": 24.88817891373802, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 31160 + }, + { + "epoch": 24.888977635782748, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 31161 + }, + { + "epoch": 24.889776357827476, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 31162 + }, + { + "epoch": 24.890575079872203, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 31163 + }, + { + "epoch": 24.891373801916934, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 31164 + }, + { + "epoch": 24.892172523961662, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 31165 + }, + { + "epoch": 24.89297124600639, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 31166 + }, + { + "epoch": 24.893769968051117, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 31167 + }, + { + "epoch": 24.894568690095845, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 31168 + }, + { + "epoch": 24.895367412140576, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 31169 + }, + { + "epoch": 24.896166134185304, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 31170 + }, + { + "epoch": 24.89696485623003, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 31171 + }, + { + "epoch": 24.89776357827476, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31172 + }, + { + "epoch": 24.89856230031949, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 31173 + }, + { + "epoch": 24.899361022364218, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 31174 + }, + { + "epoch": 24.900159744408946, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 31175 + }, + { + "epoch": 24.900958466453673, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 31176 + }, + { + "epoch": 24.9017571884984, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 31177 + }, + { + "epoch": 24.902555910543132, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31178 + }, + { + "epoch": 24.90335463258786, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 31179 + }, + { + "epoch": 24.904153354632587, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 31180 + }, + { + "epoch": 24.904952076677315, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 31181 + }, + { + "epoch": 24.905750798722046, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 31182 + }, + { + "epoch": 24.906549520766774, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 31183 + }, + { + "epoch": 24.9073482428115, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 31184 + }, + { + "epoch": 24.90814696485623, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 31185 + }, + { + "epoch": 24.908945686900957, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 31186 + }, + { + "epoch": 24.909744408945688, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 31187 + }, + { + "epoch": 24.910543130990416, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 31188 + }, + { + "epoch": 24.911341853035143, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 31189 + }, + { + "epoch": 24.91214057507987, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 31190 + }, + { + "epoch": 24.912939297124602, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 31191 + }, + { + "epoch": 24.91373801916933, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 31192 + }, + { + "epoch": 24.914536741214057, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 31193 + }, + { + "epoch": 24.915335463258785, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31194 + }, + { + "epoch": 24.916134185303516, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 31195 + }, + { + "epoch": 24.916932907348244, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 31196 + }, + { + "epoch": 24.91773162939297, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31197 + }, + { + "epoch": 24.9185303514377, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 31198 + }, + { + "epoch": 24.919329073482427, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 31199 + }, + { + "epoch": 24.920127795527158, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 31200 + }, + { + "epoch": 24.920926517571885, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 31201 + }, + { + "epoch": 24.921725239616613, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 31202 + }, + { + "epoch": 24.92252396166134, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 31203 + }, + { + "epoch": 24.923322683706072, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31204 + }, + { + "epoch": 24.9241214057508, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 31205 + }, + { + "epoch": 24.924920127795527, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 31206 + }, + { + "epoch": 24.925718849840255, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 31207 + }, + { + "epoch": 24.926517571884983, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 31208 + }, + { + "epoch": 24.927316293929714, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 31209 + }, + { + "epoch": 24.92811501597444, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 31210 + }, + { + "epoch": 24.92891373801917, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 31211 + }, + { + "epoch": 24.929712460063897, + "grad_norm": 0.0927734375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 31212 + }, + { + "epoch": 24.930511182108628, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 31213 + }, + { + "epoch": 24.931309904153355, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 31214 + }, + { + "epoch": 24.932108626198083, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 31215 + }, + { + "epoch": 24.93290734824281, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 31216 + }, + { + "epoch": 24.93370607028754, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 31217 + }, + { + "epoch": 24.93450479233227, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 31218 + }, + { + "epoch": 24.935303514376997, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 31219 + }, + { + "epoch": 24.936102236421725, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 31220 + }, + { + "epoch": 24.936900958466452, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 31221 + }, + { + "epoch": 24.937699680511184, + "grad_norm": 0.09765625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 31222 + }, + { + "epoch": 24.93849840255591, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 31223 + }, + { + "epoch": 24.93929712460064, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31224 + }, + { + "epoch": 24.940095846645367, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 31225 + }, + { + "epoch": 24.940894568690094, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31226 + }, + { + "epoch": 24.941693290734825, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 31227 + }, + { + "epoch": 24.942492012779553, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 31228 + }, + { + "epoch": 24.94329073482428, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 31229 + }, + { + "epoch": 24.94408945686901, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 31230 + }, + { + "epoch": 24.94488817891374, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 31231 + }, + { + "epoch": 24.945686900958467, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 31232 + }, + { + "epoch": 24.946485623003195, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31233 + }, + { + "epoch": 24.947284345047922, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 31234 + }, + { + "epoch": 24.94808306709265, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 31235 + }, + { + "epoch": 24.94888178913738, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 31236 + }, + { + "epoch": 24.94968051118211, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 31237 + }, + { + "epoch": 24.950479233226837, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 31238 + }, + { + "epoch": 24.951277955271564, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 31239 + }, + { + "epoch": 24.952076677316295, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 31240 + }, + { + "epoch": 24.952875399361023, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 31241 + }, + { + "epoch": 24.95367412140575, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 31242 + }, + { + "epoch": 24.95447284345048, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 31243 + }, + { + "epoch": 24.955271565495206, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 31244 + }, + { + "epoch": 24.956070287539937, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 31245 + }, + { + "epoch": 24.956869009584665, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 31246 + }, + { + "epoch": 24.957667731629392, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 31247 + }, + { + "epoch": 24.95846645367412, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 31248 + }, + { + "epoch": 24.95926517571885, + "grad_norm": 0.330078125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 31249 + }, + { + "epoch": 24.96006389776358, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 31250 + }, + { + "epoch": 24.960862619808307, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 31251 + }, + { + "epoch": 24.961661341853034, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 31252 + }, + { + "epoch": 24.962460063897765, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31253 + }, + { + "epoch": 24.963258785942493, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 31254 + }, + { + "epoch": 24.96405750798722, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 31255 + }, + { + "epoch": 24.96485623003195, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 31256 + }, + { + "epoch": 24.965654952076676, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 31257 + }, + { + "epoch": 24.966453674121407, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 31258 + }, + { + "epoch": 24.967252396166135, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 31259 + }, + { + "epoch": 24.968051118210862, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 31260 + }, + { + "epoch": 24.96884984025559, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 31261 + }, + { + "epoch": 24.96964856230032, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 31262 + }, + { + "epoch": 24.97044728434505, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 31263 + }, + { + "epoch": 24.971246006389777, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 31264 + }, + { + "epoch": 24.972044728434504, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 31265 + }, + { + "epoch": 24.972843450479232, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 31266 + }, + { + "epoch": 24.973642172523963, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31267 + }, + { + "epoch": 24.97444089456869, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 31268 + }, + { + "epoch": 24.97523961661342, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31269 + }, + { + "epoch": 24.976038338658146, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 31270 + }, + { + "epoch": 24.976837060702877, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 31271 + }, + { + "epoch": 24.977635782747605, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31272 + }, + { + "epoch": 24.978434504792332, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 31273 + }, + { + "epoch": 24.97923322683706, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 31274 + }, + { + "epoch": 24.980031948881788, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 31275 + }, + { + "epoch": 24.98083067092652, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 31276 + }, + { + "epoch": 24.981629392971247, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0746, + "step": 31277 + }, + { + "epoch": 24.982428115015974, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 31278 + }, + { + "epoch": 24.983226837060702, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31279 + }, + { + "epoch": 24.984025559105433, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 31280 + }, + { + "epoch": 24.98482428115016, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31281 + }, + { + "epoch": 24.98562300319489, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 31282 + }, + { + "epoch": 24.986421725239616, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 31283 + }, + { + "epoch": 24.987220447284344, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 31284 + }, + { + "epoch": 24.988019169329075, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 31285 + }, + { + "epoch": 24.988817891373802, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 31286 + }, + { + "epoch": 24.98961661341853, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 31287 + }, + { + "epoch": 24.990415335463258, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 31288 + }, + { + "epoch": 24.99121405750799, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 31289 + }, + { + "epoch": 24.992012779552716, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 31290 + }, + { + "epoch": 24.992811501597444, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 31291 + }, + { + "epoch": 24.99361022364217, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 31292 + }, + { + "epoch": 24.9944089456869, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 31293 + }, + { + "epoch": 24.99520766773163, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 31294 + }, + { + "epoch": 24.99600638977636, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 31295 + }, + { + "epoch": 24.996805111821086, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0736, + "step": 31296 + }, + { + "epoch": 24.997603833865814, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 31297 + }, + { + "epoch": 24.998402555910545, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 31298 + }, + { + "epoch": 24.999201277955272, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 31299 + }, + { + "epoch": 25.0, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 31300 + }, + { + "epoch": 25.000798722044728, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 31301 + }, + { + "epoch": 25.001597444089455, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 31302 + }, + { + "epoch": 25.002396166134186, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 31303 + }, + { + "epoch": 25.003194888178914, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 31304 + }, + { + "epoch": 25.00399361022364, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 31305 + }, + { + "epoch": 25.00479233226837, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 31306 + }, + { + "epoch": 25.0055910543131, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31307 + }, + { + "epoch": 25.00638977635783, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 31308 + }, + { + "epoch": 25.007188498402556, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 31309 + }, + { + "epoch": 25.007987220447284, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 31310 + }, + { + "epoch": 25.00878594249201, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 31311 + }, + { + "epoch": 25.009584664536742, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 31312 + }, + { + "epoch": 25.01038338658147, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.1039, + "step": 31313 + }, + { + "epoch": 25.011182108626198, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 31314 + }, + { + "epoch": 25.011980830670925, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 31315 + }, + { + "epoch": 25.012779552715656, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 31316 + }, + { + "epoch": 25.013578274760384, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 31317 + }, + { + "epoch": 25.01437699680511, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 31318 + }, + { + "epoch": 25.01517571884984, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 31319 + }, + { + "epoch": 25.015974440894567, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 31320 + }, + { + "epoch": 25.016773162939298, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 31321 + }, + { + "epoch": 25.017571884984026, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 31322 + }, + { + "epoch": 25.018370607028753, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 31323 + }, + { + "epoch": 25.01916932907348, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 31324 + }, + { + "epoch": 25.019968051118212, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 31325 + }, + { + "epoch": 25.02076677316294, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 31326 + }, + { + "epoch": 25.021565495207668, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 31327 + }, + { + "epoch": 25.022364217252395, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 31328 + }, + { + "epoch": 25.023162939297123, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 31329 + }, + { + "epoch": 25.023961661341854, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 31330 + }, + { + "epoch": 25.02476038338658, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 31331 + }, + { + "epoch": 25.02555910543131, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 31332 + }, + { + "epoch": 25.026357827476037, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 31333 + }, + { + "epoch": 25.027156549520768, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 31334 + }, + { + "epoch": 25.027955271565496, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 31335 + }, + { + "epoch": 25.028753993610223, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 31336 + }, + { + "epoch": 25.02955271565495, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 31337 + }, + { + "epoch": 25.03035143769968, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 31338 + }, + { + "epoch": 25.03115015974441, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 31339 + }, + { + "epoch": 25.031948881789138, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31340 + }, + { + "epoch": 25.032747603833865, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 31341 + }, + { + "epoch": 25.033546325878593, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 31342 + }, + { + "epoch": 25.034345047923324, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 31343 + }, + { + "epoch": 25.03514376996805, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 31344 + }, + { + "epoch": 25.03594249201278, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 31345 + }, + { + "epoch": 25.036741214057507, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 31346 + }, + { + "epoch": 25.037539936102238, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31347 + }, + { + "epoch": 25.038338658146966, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 31348 + }, + { + "epoch": 25.039137380191693, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 31349 + }, + { + "epoch": 25.03993610223642, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 31350 + }, + { + "epoch": 25.04073482428115, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 31351 + }, + { + "epoch": 25.04153354632588, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 31352 + }, + { + "epoch": 25.042332268370608, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 31353 + }, + { + "epoch": 25.043130990415335, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 31354 + }, + { + "epoch": 25.043929712460063, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 31355 + }, + { + "epoch": 25.044728434504794, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 31356 + }, + { + "epoch": 25.04552715654952, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 31357 + }, + { + "epoch": 25.04632587859425, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 31358 + }, + { + "epoch": 25.047124600638977, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31359 + }, + { + "epoch": 25.047923322683705, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 31360 + }, + { + "epoch": 25.048722044728436, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 31361 + }, + { + "epoch": 25.049520766773163, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 31362 + }, + { + "epoch": 25.05031948881789, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31363 + }, + { + "epoch": 25.05111821086262, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 31364 + }, + { + "epoch": 25.05191693290735, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 31365 + }, + { + "epoch": 25.052715654952078, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 31366 + }, + { + "epoch": 25.053514376996805, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31367 + }, + { + "epoch": 25.054313099041533, + "grad_norm": 0.09521484375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 31368 + }, + { + "epoch": 25.05511182108626, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 31369 + }, + { + "epoch": 25.05591054313099, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 31370 + }, + { + "epoch": 25.05670926517572, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 31371 + }, + { + "epoch": 25.057507987220447, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 31372 + }, + { + "epoch": 25.058306709265175, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 31373 + }, + { + "epoch": 25.059105431309906, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 31374 + }, + { + "epoch": 25.059904153354633, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 31375 + }, + { + "epoch": 25.06070287539936, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 31376 + }, + { + "epoch": 25.06150159744409, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31377 + }, + { + "epoch": 25.062300319488816, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 31378 + }, + { + "epoch": 25.063099041533548, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 31379 + }, + { + "epoch": 25.063897763578275, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 31380 + }, + { + "epoch": 25.064696485623003, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 31381 + }, + { + "epoch": 25.06549520766773, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 31382 + }, + { + "epoch": 25.06629392971246, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 31383 + }, + { + "epoch": 25.06709265175719, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 31384 + }, + { + "epoch": 25.067891373801917, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 31385 + }, + { + "epoch": 25.068690095846645, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 31386 + }, + { + "epoch": 25.069488817891372, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 31387 + }, + { + "epoch": 25.070287539936103, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 31388 + }, + { + "epoch": 25.07108626198083, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 31389 + }, + { + "epoch": 25.07188498402556, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 31390 + }, + { + "epoch": 25.072683706070286, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 31391 + }, + { + "epoch": 25.073482428115017, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 31392 + }, + { + "epoch": 25.074281150159745, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 31393 + }, + { + "epoch": 25.075079872204473, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 31394 + }, + { + "epoch": 25.0758785942492, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 31395 + }, + { + "epoch": 25.076677316293928, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 31396 + }, + { + "epoch": 25.07747603833866, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1046, + "step": 31397 + }, + { + "epoch": 25.078274760383387, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 31398 + }, + { + "epoch": 25.079073482428115, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 31399 + }, + { + "epoch": 25.079872204472842, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 31400 + }, + { + "epoch": 25.080670926517573, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 31401 + }, + { + "epoch": 25.0814696485623, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 31402 + }, + { + "epoch": 25.08226837060703, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 31403 + }, + { + "epoch": 25.083067092651756, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 31404 + }, + { + "epoch": 25.083865814696484, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 31405 + }, + { + "epoch": 25.084664536741215, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 31406 + }, + { + "epoch": 25.085463258785943, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 31407 + }, + { + "epoch": 25.08626198083067, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 31408 + }, + { + "epoch": 25.087060702875398, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 31409 + }, + { + "epoch": 25.08785942492013, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 31410 + }, + { + "epoch": 25.088658146964857, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 31411 + }, + { + "epoch": 25.089456869009584, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 31412 + }, + { + "epoch": 25.090255591054312, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 31413 + }, + { + "epoch": 25.09105431309904, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 31414 + }, + { + "epoch": 25.09185303514377, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 31415 + }, + { + "epoch": 25.0926517571885, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 31416 + }, + { + "epoch": 25.093450479233226, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 31417 + }, + { + "epoch": 25.094249201277954, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 31418 + }, + { + "epoch": 25.095047923322685, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 31419 + }, + { + "epoch": 25.095846645367413, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31420 + }, + { + "epoch": 25.09664536741214, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 31421 + }, + { + "epoch": 25.097444089456868, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 31422 + }, + { + "epoch": 25.0982428115016, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 31423 + }, + { + "epoch": 25.099041533546327, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 31424 + }, + { + "epoch": 25.099840255591054, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 31425 + }, + { + "epoch": 25.100638977635782, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 31426 + }, + { + "epoch": 25.10143769968051, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 31427 + }, + { + "epoch": 25.10223642172524, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 31428 + }, + { + "epoch": 25.10303514376997, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 31429 + }, + { + "epoch": 25.103833865814696, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 31430 + }, + { + "epoch": 25.104632587859424, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 31431 + }, + { + "epoch": 25.105431309904155, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 31432 + }, + { + "epoch": 25.106230031948883, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 31433 + }, + { + "epoch": 25.10702875399361, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 31434 + }, + { + "epoch": 25.107827476038338, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 31435 + }, + { + "epoch": 25.108626198083066, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1041, + "step": 31436 + }, + { + "epoch": 25.109424920127797, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 31437 + }, + { + "epoch": 25.110223642172524, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 31438 + }, + { + "epoch": 25.111022364217252, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 31439 + }, + { + "epoch": 25.11182108626198, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 31440 + }, + { + "epoch": 25.11261980830671, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 31441 + }, + { + "epoch": 25.11341853035144, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 31442 + }, + { + "epoch": 25.114217252396166, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 31443 + }, + { + "epoch": 25.115015974440894, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 31444 + }, + { + "epoch": 25.11581469648562, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 31445 + }, + { + "epoch": 25.116613418530353, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31446 + }, + { + "epoch": 25.11741214057508, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31447 + }, + { + "epoch": 25.118210862619808, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 31448 + }, + { + "epoch": 25.119009584664536, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 31449 + }, + { + "epoch": 25.119808306709267, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 31450 + }, + { + "epoch": 25.120607028753994, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 31451 + }, + { + "epoch": 25.121405750798722, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 31452 + }, + { + "epoch": 25.12220447284345, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 31453 + }, + { + "epoch": 25.123003194888177, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 31454 + }, + { + "epoch": 25.12380191693291, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 31455 + }, + { + "epoch": 25.124600638977636, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 31456 + }, + { + "epoch": 25.125399361022364, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 31457 + }, + { + "epoch": 25.12619808306709, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 31458 + }, + { + "epoch": 25.126996805111823, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 31459 + }, + { + "epoch": 25.12779552715655, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 31460 + }, + { + "epoch": 25.128594249201278, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 31461 + }, + { + "epoch": 25.129392971246006, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 31462 + }, + { + "epoch": 25.130191693290733, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 31463 + }, + { + "epoch": 25.130990415335464, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 31464 + }, + { + "epoch": 25.131789137380192, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 31465 + }, + { + "epoch": 25.13258785942492, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 31466 + }, + { + "epoch": 25.133386581469647, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 31467 + }, + { + "epoch": 25.13418530351438, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31468 + }, + { + "epoch": 25.134984025559106, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 31469 + }, + { + "epoch": 25.135782747603834, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 31470 + }, + { + "epoch": 25.13658146964856, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 31471 + }, + { + "epoch": 25.13738019169329, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 31472 + }, + { + "epoch": 25.13817891373802, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31473 + }, + { + "epoch": 25.138977635782748, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 31474 + }, + { + "epoch": 25.139776357827476, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 31475 + }, + { + "epoch": 25.140575079872203, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31476 + }, + { + "epoch": 25.141373801916934, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 31477 + }, + { + "epoch": 25.142172523961662, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 31478 + }, + { + "epoch": 25.14297124600639, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 31479 + }, + { + "epoch": 25.143769968051117, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 31480 + }, + { + "epoch": 25.144568690095845, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 31481 + }, + { + "epoch": 25.145367412140576, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 31482 + }, + { + "epoch": 25.146166134185304, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 31483 + }, + { + "epoch": 25.14696485623003, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 31484 + }, + { + "epoch": 25.14776357827476, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 31485 + }, + { + "epoch": 25.14856230031949, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31486 + }, + { + "epoch": 25.149361022364218, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 31487 + }, + { + "epoch": 25.150159744408946, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31488 + }, + { + "epoch": 25.150958466453673, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 31489 + }, + { + "epoch": 25.1517571884984, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 31490 + }, + { + "epoch": 25.152555910543132, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 31491 + }, + { + "epoch": 25.15335463258786, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 31492 + }, + { + "epoch": 25.154153354632587, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 31493 + }, + { + "epoch": 25.154952076677315, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31494 + }, + { + "epoch": 25.155750798722046, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 31495 + }, + { + "epoch": 25.156549520766774, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 31496 + }, + { + "epoch": 25.1573482428115, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 31497 + }, + { + "epoch": 25.15814696485623, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 31498 + }, + { + "epoch": 25.15894568690096, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31499 + }, + { + "epoch": 25.159744408945688, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 31500 + }, + { + "epoch": 25.160543130990416, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 31501 + }, + { + "epoch": 25.161341853035143, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 31502 + }, + { + "epoch": 25.16214057507987, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31503 + }, + { + "epoch": 25.162939297124602, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 31504 + }, + { + "epoch": 25.16373801916933, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 31505 + }, + { + "epoch": 25.164536741214057, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 31506 + }, + { + "epoch": 25.165335463258785, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 31507 + }, + { + "epoch": 25.166134185303516, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 31508 + }, + { + "epoch": 25.166932907348244, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31509 + }, + { + "epoch": 25.16773162939297, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 31510 + }, + { + "epoch": 25.1685303514377, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 31511 + }, + { + "epoch": 25.169329073482427, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 31512 + }, + { + "epoch": 25.170127795527158, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31513 + }, + { + "epoch": 25.170926517571885, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 31514 + }, + { + "epoch": 25.171725239616613, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 31515 + }, + { + "epoch": 25.17252396166134, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 31516 + }, + { + "epoch": 25.173322683706072, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 31517 + }, + { + "epoch": 25.1741214057508, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 31518 + }, + { + "epoch": 25.174920127795527, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 31519 + }, + { + "epoch": 25.175718849840255, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 31520 + }, + { + "epoch": 25.176517571884983, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 31521 + }, + { + "epoch": 25.177316293929714, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 31522 + }, + { + "epoch": 25.17811501597444, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 31523 + }, + { + "epoch": 25.17891373801917, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31524 + }, + { + "epoch": 25.179712460063897, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 31525 + }, + { + "epoch": 25.180511182108628, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 31526 + }, + { + "epoch": 25.181309904153355, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 31527 + }, + { + "epoch": 25.182108626198083, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 31528 + }, + { + "epoch": 25.18290734824281, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 31529 + }, + { + "epoch": 25.18370607028754, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 31530 + }, + { + "epoch": 25.18450479233227, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 31531 + }, + { + "epoch": 25.185303514376997, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 31532 + }, + { + "epoch": 25.186102236421725, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 31533 + }, + { + "epoch": 25.186900958466452, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 31534 + }, + { + "epoch": 25.187699680511184, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 31535 + }, + { + "epoch": 25.18849840255591, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 31536 + }, + { + "epoch": 25.18929712460064, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 31537 + }, + { + "epoch": 25.190095846645367, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 31538 + }, + { + "epoch": 25.190894568690094, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 31539 + }, + { + "epoch": 25.191693290734825, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 31540 + }, + { + "epoch": 25.192492012779553, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 31541 + }, + { + "epoch": 25.19329073482428, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 31542 + }, + { + "epoch": 25.19408945686901, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 31543 + }, + { + "epoch": 25.19488817891374, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 31544 + }, + { + "epoch": 25.195686900958467, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 31545 + }, + { + "epoch": 25.196485623003195, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 31546 + }, + { + "epoch": 25.197284345047922, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31547 + }, + { + "epoch": 25.19808306709265, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 31548 + }, + { + "epoch": 25.19888178913738, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 31549 + }, + { + "epoch": 25.19968051118211, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 31550 + }, + { + "epoch": 25.200479233226837, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 31551 + }, + { + "epoch": 25.201277955271564, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 31552 + }, + { + "epoch": 25.202076677316295, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 31553 + }, + { + "epoch": 25.202875399361023, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 31554 + }, + { + "epoch": 25.20367412140575, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31555 + }, + { + "epoch": 25.20447284345048, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 31556 + }, + { + "epoch": 25.205271565495206, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 31557 + }, + { + "epoch": 25.206070287539937, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 31558 + }, + { + "epoch": 25.206869009584665, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 31559 + }, + { + "epoch": 25.207667731629392, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 31560 + }, + { + "epoch": 25.20846645367412, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 31561 + }, + { + "epoch": 25.20926517571885, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 31562 + }, + { + "epoch": 25.21006389776358, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31563 + }, + { + "epoch": 25.210862619808307, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.1026, + "step": 31564 + }, + { + "epoch": 25.211661341853034, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 31565 + }, + { + "epoch": 25.212460063897762, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 31566 + }, + { + "epoch": 25.213258785942493, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 31567 + }, + { + "epoch": 25.21405750798722, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 31568 + }, + { + "epoch": 25.21485623003195, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31569 + }, + { + "epoch": 25.215654952076676, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 31570 + }, + { + "epoch": 25.216453674121407, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 31571 + }, + { + "epoch": 25.217252396166135, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 31572 + }, + { + "epoch": 25.218051118210862, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 31573 + }, + { + "epoch": 25.21884984025559, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 31574 + }, + { + "epoch": 25.21964856230032, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 31575 + }, + { + "epoch": 25.22044728434505, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 31576 + }, + { + "epoch": 25.221246006389777, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 31577 + }, + { + "epoch": 25.222044728434504, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 31578 + }, + { + "epoch": 25.222843450479232, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 31579 + }, + { + "epoch": 25.223642172523963, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 31580 + }, + { + "epoch": 25.22444089456869, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 31581 + }, + { + "epoch": 25.22523961661342, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 31582 + }, + { + "epoch": 25.226038338658146, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31583 + }, + { + "epoch": 25.226837060702877, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 31584 + }, + { + "epoch": 25.227635782747605, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 31585 + }, + { + "epoch": 25.228434504792332, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 31586 + }, + { + "epoch": 25.22923322683706, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 31587 + }, + { + "epoch": 25.230031948881788, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 31588 + }, + { + "epoch": 25.23083067092652, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 31589 + }, + { + "epoch": 25.231629392971247, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 31590 + }, + { + "epoch": 25.232428115015974, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 31591 + }, + { + "epoch": 25.233226837060702, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 31592 + }, + { + "epoch": 25.234025559105433, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 31593 + }, + { + "epoch": 25.23482428115016, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 31594 + }, + { + "epoch": 25.23562300319489, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 31595 + }, + { + "epoch": 25.236421725239616, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 31596 + }, + { + "epoch": 25.237220447284344, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 31597 + }, + { + "epoch": 25.238019169329075, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 31598 + }, + { + "epoch": 25.238817891373802, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 31599 + }, + { + "epoch": 25.23961661341853, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 31600 + }, + { + "epoch": 25.240415335463258, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 31601 + }, + { + "epoch": 25.24121405750799, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 31602 + }, + { + "epoch": 25.242012779552716, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 31603 + }, + { + "epoch": 25.242811501597444, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 31604 + }, + { + "epoch": 25.24361022364217, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 31605 + }, + { + "epoch": 25.2444089456869, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 31606 + }, + { + "epoch": 25.24520766773163, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 31607 + }, + { + "epoch": 25.24600638977636, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 31608 + }, + { + "epoch": 25.246805111821086, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 31609 + }, + { + "epoch": 25.247603833865814, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 31610 + }, + { + "epoch": 25.248402555910545, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 31611 + }, + { + "epoch": 25.249201277955272, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 31612 + }, + { + "epoch": 25.25, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 31613 + }, + { + "epoch": 25.250798722044728, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 31614 + }, + { + "epoch": 25.251597444089455, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 31615 + }, + { + "epoch": 25.252396166134186, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 31616 + }, + { + "epoch": 25.253194888178914, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 31617 + }, + { + "epoch": 25.25399361022364, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 31618 + }, + { + "epoch": 25.25479233226837, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 31619 + }, + { + "epoch": 25.2555910543131, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 31620 + }, + { + "epoch": 25.25638977635783, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 31621 + }, + { + "epoch": 25.257188498402556, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31622 + }, + { + "epoch": 25.257987220447284, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 31623 + }, + { + "epoch": 25.25878594249201, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 31624 + }, + { + "epoch": 25.259584664536742, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 31625 + }, + { + "epoch": 25.26038338658147, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 31626 + }, + { + "epoch": 25.261182108626198, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31627 + }, + { + "epoch": 25.261980830670925, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 31628 + }, + { + "epoch": 25.262779552715656, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 31629 + }, + { + "epoch": 25.263578274760384, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 31630 + }, + { + "epoch": 25.26437699680511, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 31631 + }, + { + "epoch": 25.26517571884984, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 31632 + }, + { + "epoch": 25.265974440894567, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 31633 + }, + { + "epoch": 25.266773162939298, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 31634 + }, + { + "epoch": 25.267571884984026, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 31635 + }, + { + "epoch": 25.268370607028753, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 31636 + }, + { + "epoch": 25.26916932907348, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 31637 + }, + { + "epoch": 25.269968051118212, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 31638 + }, + { + "epoch": 25.27076677316294, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 31639 + }, + { + "epoch": 25.271565495207668, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 31640 + }, + { + "epoch": 25.272364217252395, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 31641 + }, + { + "epoch": 25.273162939297123, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 31642 + }, + { + "epoch": 25.273961661341854, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31643 + }, + { + "epoch": 25.27476038338658, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 31644 + }, + { + "epoch": 25.27555910543131, + "grad_norm": 0.09619140625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 31645 + }, + { + "epoch": 25.276357827476037, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 31646 + }, + { + "epoch": 25.277156549520768, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 31647 + }, + { + "epoch": 25.277955271565496, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31648 + }, + { + "epoch": 25.278753993610223, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 31649 + }, + { + "epoch": 25.27955271565495, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 31650 + }, + { + "epoch": 25.28035143769968, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 31651 + }, + { + "epoch": 25.28115015974441, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 31652 + }, + { + "epoch": 25.281948881789138, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 31653 + }, + { + "epoch": 25.282747603833865, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 31654 + }, + { + "epoch": 25.283546325878593, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 31655 + }, + { + "epoch": 25.284345047923324, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 31656 + }, + { + "epoch": 25.28514376996805, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 31657 + }, + { + "epoch": 25.28594249201278, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 31658 + }, + { + "epoch": 25.286741214057507, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 31659 + }, + { + "epoch": 25.287539936102238, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 31660 + }, + { + "epoch": 25.288338658146966, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 31661 + }, + { + "epoch": 25.289137380191693, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0748, + "step": 31662 + }, + { + "epoch": 25.28993610223642, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 31663 + }, + { + "epoch": 25.29073482428115, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 31664 + }, + { + "epoch": 25.29153354632588, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 31665 + }, + { + "epoch": 25.292332268370608, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 31666 + }, + { + "epoch": 25.293130990415335, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 31667 + }, + { + "epoch": 25.293929712460063, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 31668 + }, + { + "epoch": 25.294728434504794, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31669 + }, + { + "epoch": 25.29552715654952, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 31670 + }, + { + "epoch": 25.29632587859425, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 31671 + }, + { + "epoch": 25.297124600638977, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 31672 + }, + { + "epoch": 25.297923322683705, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 31673 + }, + { + "epoch": 25.298722044728436, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 31674 + }, + { + "epoch": 25.299520766773163, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 31675 + }, + { + "epoch": 25.30031948881789, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 31676 + }, + { + "epoch": 25.30111821086262, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 31677 + }, + { + "epoch": 25.30191693290735, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 31678 + }, + { + "epoch": 25.302715654952078, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31679 + }, + { + "epoch": 25.303514376996805, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 31680 + }, + { + "epoch": 25.304313099041533, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 31681 + }, + { + "epoch": 25.30511182108626, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 31682 + }, + { + "epoch": 25.30591054313099, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31683 + }, + { + "epoch": 25.30670926517572, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 31684 + }, + { + "epoch": 25.307507987220447, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 31685 + }, + { + "epoch": 25.308306709265175, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 31686 + }, + { + "epoch": 25.309105431309906, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 31687 + }, + { + "epoch": 25.309904153354633, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 31688 + }, + { + "epoch": 25.31070287539936, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 31689 + }, + { + "epoch": 25.31150159744409, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 31690 + }, + { + "epoch": 25.312300319488816, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 31691 + }, + { + "epoch": 25.313099041533548, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 31692 + }, + { + "epoch": 25.313897763578275, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 31693 + }, + { + "epoch": 25.314696485623003, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 31694 + }, + { + "epoch": 25.31549520766773, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 31695 + }, + { + "epoch": 25.31629392971246, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 31696 + }, + { + "epoch": 25.31709265175719, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 31697 + }, + { + "epoch": 25.317891373801917, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 31698 + }, + { + "epoch": 25.318690095846645, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 31699 + }, + { + "epoch": 25.319488817891372, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 31700 + }, + { + "epoch": 25.320287539936103, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 31701 + }, + { + "epoch": 25.32108626198083, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 31702 + }, + { + "epoch": 25.32188498402556, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 31703 + }, + { + "epoch": 25.322683706070286, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 31704 + }, + { + "epoch": 25.323482428115017, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 31705 + }, + { + "epoch": 25.324281150159745, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 31706 + }, + { + "epoch": 25.325079872204473, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 31707 + }, + { + "epoch": 25.3258785942492, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 31708 + }, + { + "epoch": 25.326677316293928, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 31709 + }, + { + "epoch": 25.32747603833866, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 31710 + }, + { + "epoch": 25.328274760383387, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31711 + }, + { + "epoch": 25.329073482428115, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 31712 + }, + { + "epoch": 25.329872204472842, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 31713 + }, + { + "epoch": 25.330670926517573, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 31714 + }, + { + "epoch": 25.3314696485623, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 31715 + }, + { + "epoch": 25.33226837060703, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 31716 + }, + { + "epoch": 25.333067092651756, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 31717 + }, + { + "epoch": 25.333865814696484, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 31718 + }, + { + "epoch": 25.334664536741215, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 31719 + }, + { + "epoch": 25.335463258785943, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 31720 + }, + { + "epoch": 25.33626198083067, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 31721 + }, + { + "epoch": 25.337060702875398, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 31722 + }, + { + "epoch": 25.33785942492013, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 31723 + }, + { + "epoch": 25.338658146964857, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 31724 + }, + { + "epoch": 25.339456869009584, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 31725 + }, + { + "epoch": 25.340255591054312, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 31726 + }, + { + "epoch": 25.341054313099043, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 31727 + }, + { + "epoch": 25.34185303514377, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 31728 + }, + { + "epoch": 25.3426517571885, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31729 + }, + { + "epoch": 25.343450479233226, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31730 + }, + { + "epoch": 25.344249201277954, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 31731 + }, + { + "epoch": 25.345047923322685, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 31732 + }, + { + "epoch": 25.345846645367413, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 31733 + }, + { + "epoch": 25.34664536741214, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 31734 + }, + { + "epoch": 25.347444089456868, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 31735 + }, + { + "epoch": 25.3482428115016, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 31736 + }, + { + "epoch": 25.349041533546327, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 31737 + }, + { + "epoch": 25.349840255591054, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 31738 + }, + { + "epoch": 25.350638977635782, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 31739 + }, + { + "epoch": 25.35143769968051, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 31740 + }, + { + "epoch": 25.35223642172524, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 31741 + }, + { + "epoch": 25.35303514376997, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 31742 + }, + { + "epoch": 25.353833865814696, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 31743 + }, + { + "epoch": 25.354632587859424, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 31744 + }, + { + "epoch": 25.355431309904155, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 31745 + }, + { + "epoch": 25.356230031948883, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 31746 + }, + { + "epoch": 25.35702875399361, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 31747 + }, + { + "epoch": 25.357827476038338, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 31748 + }, + { + "epoch": 25.358626198083066, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 31749 + }, + { + "epoch": 25.359424920127797, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 31750 + }, + { + "epoch": 25.360223642172524, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 31751 + }, + { + "epoch": 25.361022364217252, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 31752 + }, + { + "epoch": 25.36182108626198, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 31753 + }, + { + "epoch": 25.36261980830671, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 31754 + }, + { + "epoch": 25.36341853035144, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 31755 + }, + { + "epoch": 25.364217252396166, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 31756 + }, + { + "epoch": 25.365015974440894, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 31757 + }, + { + "epoch": 25.36581469648562, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 31758 + }, + { + "epoch": 25.366613418530353, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 31759 + }, + { + "epoch": 25.36741214057508, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 31760 + }, + { + "epoch": 25.368210862619808, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 31761 + }, + { + "epoch": 25.369009584664536, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 31762 + }, + { + "epoch": 25.369808306709267, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 31763 + }, + { + "epoch": 25.370607028753994, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31764 + }, + { + "epoch": 25.371405750798722, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 31765 + }, + { + "epoch": 25.37220447284345, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 31766 + }, + { + "epoch": 25.373003194888177, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 31767 + }, + { + "epoch": 25.37380191693291, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 31768 + }, + { + "epoch": 25.374600638977636, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 31769 + }, + { + "epoch": 25.375399361022364, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 31770 + }, + { + "epoch": 25.37619808306709, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 31771 + }, + { + "epoch": 25.376996805111823, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 31772 + }, + { + "epoch": 25.37779552715655, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 31773 + }, + { + "epoch": 25.378594249201278, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 31774 + }, + { + "epoch": 25.379392971246006, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 31775 + }, + { + "epoch": 25.380191693290733, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 31776 + }, + { + "epoch": 25.380990415335464, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 31777 + }, + { + "epoch": 25.381789137380192, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 31778 + }, + { + "epoch": 25.38258785942492, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 31779 + }, + { + "epoch": 25.383386581469647, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 31780 + }, + { + "epoch": 25.38418530351438, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 31781 + }, + { + "epoch": 25.384984025559106, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 31782 + }, + { + "epoch": 25.385782747603834, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 31783 + }, + { + "epoch": 25.38658146964856, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 31784 + }, + { + "epoch": 25.38738019169329, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 31785 + }, + { + "epoch": 25.38817891373802, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 31786 + }, + { + "epoch": 25.388977635782748, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 31787 + }, + { + "epoch": 25.389776357827476, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 31788 + }, + { + "epoch": 25.390575079872203, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 31789 + }, + { + "epoch": 25.391373801916934, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 31790 + }, + { + "epoch": 25.392172523961662, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 31791 + }, + { + "epoch": 25.39297124600639, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31792 + }, + { + "epoch": 25.393769968051117, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 31793 + }, + { + "epoch": 25.394568690095845, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 31794 + }, + { + "epoch": 25.395367412140576, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 31795 + }, + { + "epoch": 25.396166134185304, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 31796 + }, + { + "epoch": 25.39696485623003, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 31797 + }, + { + "epoch": 25.39776357827476, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31798 + }, + { + "epoch": 25.39856230031949, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 31799 + }, + { + "epoch": 25.399361022364218, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 31800 + }, + { + "epoch": 25.400159744408946, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31801 + }, + { + "epoch": 25.400958466453673, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 31802 + }, + { + "epoch": 25.4017571884984, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 31803 + }, + { + "epoch": 25.402555910543132, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 31804 + }, + { + "epoch": 25.40335463258786, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 31805 + }, + { + "epoch": 25.404153354632587, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 31806 + }, + { + "epoch": 25.404952076677315, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 31807 + }, + { + "epoch": 25.405750798722046, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 31808 + }, + { + "epoch": 25.406549520766774, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 31809 + }, + { + "epoch": 25.4073482428115, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 31810 + }, + { + "epoch": 25.40814696485623, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 31811 + }, + { + "epoch": 25.408945686900957, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 31812 + }, + { + "epoch": 25.409744408945688, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 31813 + }, + { + "epoch": 25.410543130990416, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 31814 + }, + { + "epoch": 25.411341853035143, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31815 + }, + { + "epoch": 25.41214057507987, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 31816 + }, + { + "epoch": 25.412939297124602, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 31817 + }, + { + "epoch": 25.41373801916933, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 31818 + }, + { + "epoch": 25.414536741214057, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 31819 + }, + { + "epoch": 25.415335463258785, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 31820 + }, + { + "epoch": 25.416134185303516, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31821 + }, + { + "epoch": 25.416932907348244, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 31822 + }, + { + "epoch": 25.41773162939297, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 31823 + }, + { + "epoch": 25.4185303514377, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 31824 + }, + { + "epoch": 25.419329073482427, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 31825 + }, + { + "epoch": 25.420127795527158, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 31826 + }, + { + "epoch": 25.420926517571885, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 31827 + }, + { + "epoch": 25.421725239616613, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 31828 + }, + { + "epoch": 25.42252396166134, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 31829 + }, + { + "epoch": 25.423322683706072, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31830 + }, + { + "epoch": 25.4241214057508, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 31831 + }, + { + "epoch": 25.424920127795527, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 31832 + }, + { + "epoch": 25.425718849840255, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 31833 + }, + { + "epoch": 25.426517571884983, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 31834 + }, + { + "epoch": 25.427316293929714, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 31835 + }, + { + "epoch": 25.42811501597444, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 31836 + }, + { + "epoch": 25.42891373801917, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 31837 + }, + { + "epoch": 25.429712460063897, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 31838 + }, + { + "epoch": 25.430511182108628, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 31839 + }, + { + "epoch": 25.431309904153355, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31840 + }, + { + "epoch": 25.432108626198083, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 31841 + }, + { + "epoch": 25.43290734824281, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 31842 + }, + { + "epoch": 25.43370607028754, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 31843 + }, + { + "epoch": 25.43450479233227, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 31844 + }, + { + "epoch": 25.435303514376997, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 31845 + }, + { + "epoch": 25.436102236421725, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 31846 + }, + { + "epoch": 25.436900958466452, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 31847 + }, + { + "epoch": 25.437699680511184, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 31848 + }, + { + "epoch": 25.43849840255591, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 31849 + }, + { + "epoch": 25.43929712460064, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 31850 + }, + { + "epoch": 25.440095846645367, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 31851 + }, + { + "epoch": 25.440894568690094, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31852 + }, + { + "epoch": 25.441693290734825, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 31853 + }, + { + "epoch": 25.442492012779553, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 31854 + }, + { + "epoch": 25.44329073482428, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 31855 + }, + { + "epoch": 25.44408945686901, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 31856 + }, + { + "epoch": 25.44488817891374, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 31857 + }, + { + "epoch": 25.445686900958467, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 31858 + }, + { + "epoch": 25.446485623003195, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 31859 + }, + { + "epoch": 25.447284345047922, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 31860 + }, + { + "epoch": 25.44808306709265, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 31861 + }, + { + "epoch": 25.44888178913738, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 31862 + }, + { + "epoch": 25.44968051118211, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 31863 + }, + { + "epoch": 25.450479233226837, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 31864 + }, + { + "epoch": 25.451277955271564, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 31865 + }, + { + "epoch": 25.452076677316295, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 31866 + }, + { + "epoch": 25.452875399361023, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 31867 + }, + { + "epoch": 25.45367412140575, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31868 + }, + { + "epoch": 25.45447284345048, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 31869 + }, + { + "epoch": 25.455271565495206, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31870 + }, + { + "epoch": 25.456070287539937, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 31871 + }, + { + "epoch": 25.456869009584665, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 31872 + }, + { + "epoch": 25.457667731629392, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 31873 + }, + { + "epoch": 25.45846645367412, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 31874 + }, + { + "epoch": 25.45926517571885, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 31875 + }, + { + "epoch": 25.46006389776358, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 31876 + }, + { + "epoch": 25.460862619808307, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 31877 + }, + { + "epoch": 25.461661341853034, + "grad_norm": 0.09326171875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 31878 + }, + { + "epoch": 25.462460063897762, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 31879 + }, + { + "epoch": 25.463258785942493, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 31880 + }, + { + "epoch": 25.46405750798722, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 31881 + }, + { + "epoch": 25.46485623003195, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 31882 + }, + { + "epoch": 25.465654952076676, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 31883 + }, + { + "epoch": 25.466453674121407, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 31884 + }, + { + "epoch": 25.467252396166135, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 31885 + }, + { + "epoch": 25.468051118210862, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 31886 + }, + { + "epoch": 25.46884984025559, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 31887 + }, + { + "epoch": 25.46964856230032, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 31888 + }, + { + "epoch": 25.47044728434505, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 31889 + }, + { + "epoch": 25.471246006389777, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 31890 + }, + { + "epoch": 25.472044728434504, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 31891 + }, + { + "epoch": 25.472843450479232, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 31892 + }, + { + "epoch": 25.473642172523963, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 31893 + }, + { + "epoch": 25.47444089456869, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31894 + }, + { + "epoch": 25.47523961661342, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 31895 + }, + { + "epoch": 25.476038338658146, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 31896 + }, + { + "epoch": 25.476837060702877, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 31897 + }, + { + "epoch": 25.477635782747605, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 31898 + }, + { + "epoch": 25.478434504792332, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31899 + }, + { + "epoch": 25.47923322683706, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 31900 + }, + { + "epoch": 25.480031948881788, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 31901 + }, + { + "epoch": 25.48083067092652, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 31902 + }, + { + "epoch": 25.481629392971247, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 31903 + }, + { + "epoch": 25.482428115015974, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 31904 + }, + { + "epoch": 25.483226837060702, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 31905 + }, + { + "epoch": 25.484025559105433, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 31906 + }, + { + "epoch": 25.48482428115016, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31907 + }, + { + "epoch": 25.48562300319489, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 31908 + }, + { + "epoch": 25.486421725239616, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 31909 + }, + { + "epoch": 25.487220447284344, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 31910 + }, + { + "epoch": 25.488019169329075, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 31911 + }, + { + "epoch": 25.488817891373802, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 31912 + }, + { + "epoch": 25.48961661341853, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 31913 + }, + { + "epoch": 25.490415335463258, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 31914 + }, + { + "epoch": 25.49121405750799, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 31915 + }, + { + "epoch": 25.492012779552716, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 31916 + }, + { + "epoch": 25.492811501597444, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 31917 + }, + { + "epoch": 25.49361022364217, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 31918 + }, + { + "epoch": 25.4944089456869, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0765, + "step": 31919 + }, + { + "epoch": 25.49520766773163, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 31920 + }, + { + "epoch": 25.49600638977636, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31921 + }, + { + "epoch": 25.496805111821086, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 31922 + }, + { + "epoch": 25.497603833865814, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 31923 + }, + { + "epoch": 25.498402555910545, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 31924 + }, + { + "epoch": 25.499201277955272, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 31925 + }, + { + "epoch": 25.5, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 31926 + }, + { + "epoch": 25.500798722044728, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 31927 + }, + { + "epoch": 25.501597444089455, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 31928 + }, + { + "epoch": 25.502396166134186, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 31929 + }, + { + "epoch": 25.503194888178914, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 31930 + }, + { + "epoch": 25.50399361022364, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 31931 + }, + { + "epoch": 25.50479233226837, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 31932 + }, + { + "epoch": 25.5055910543131, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 31933 + }, + { + "epoch": 25.50638977635783, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 31934 + }, + { + "epoch": 25.507188498402556, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 31935 + }, + { + "epoch": 25.507987220447284, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 31936 + }, + { + "epoch": 25.50878594249201, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 31937 + }, + { + "epoch": 25.509584664536742, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 31938 + }, + { + "epoch": 25.51038338658147, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.103, + "step": 31939 + }, + { + "epoch": 25.511182108626198, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 31940 + }, + { + "epoch": 25.511980830670925, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 31941 + }, + { + "epoch": 25.512779552715656, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 31942 + }, + { + "epoch": 25.513578274760384, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 31943 + }, + { + "epoch": 25.51437699680511, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 31944 + }, + { + "epoch": 25.51517571884984, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 31945 + }, + { + "epoch": 25.515974440894567, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 31946 + }, + { + "epoch": 25.516773162939298, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 31947 + }, + { + "epoch": 25.517571884984026, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 31948 + }, + { + "epoch": 25.518370607028753, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 31949 + }, + { + "epoch": 25.51916932907348, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 31950 + }, + { + "epoch": 25.519968051118212, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 31951 + }, + { + "epoch": 25.52076677316294, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 31952 + }, + { + "epoch": 25.521565495207668, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 31953 + }, + { + "epoch": 25.522364217252395, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 31954 + }, + { + "epoch": 25.523162939297123, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 31955 + }, + { + "epoch": 25.523961661341854, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 31956 + }, + { + "epoch": 25.52476038338658, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 31957 + }, + { + "epoch": 25.52555910543131, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 31958 + }, + { + "epoch": 25.526357827476037, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 31959 + }, + { + "epoch": 25.527156549520768, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 31960 + }, + { + "epoch": 25.527955271565496, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 31961 + }, + { + "epoch": 25.528753993610223, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 31962 + }, + { + "epoch": 25.52955271565495, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 31963 + }, + { + "epoch": 25.53035143769968, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 31964 + }, + { + "epoch": 25.53115015974441, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 31965 + }, + { + "epoch": 25.531948881789138, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 31966 + }, + { + "epoch": 25.532747603833865, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 31967 + }, + { + "epoch": 25.533546325878593, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 31968 + }, + { + "epoch": 25.534345047923324, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 31969 + }, + { + "epoch": 25.53514376996805, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 31970 + }, + { + "epoch": 25.53594249201278, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 31971 + }, + { + "epoch": 25.536741214057507, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 31972 + }, + { + "epoch": 25.537539936102235, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 31973 + }, + { + "epoch": 25.538338658146966, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 31974 + }, + { + "epoch": 25.539137380191693, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 31975 + }, + { + "epoch": 25.53993610223642, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 31976 + }, + { + "epoch": 25.54073482428115, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 31977 + }, + { + "epoch": 25.54153354632588, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.075, + "step": 31978 + }, + { + "epoch": 25.542332268370608, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 31979 + }, + { + "epoch": 25.543130990415335, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31980 + }, + { + "epoch": 25.543929712460063, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.106, + "step": 31981 + }, + { + "epoch": 25.544728434504794, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 31982 + }, + { + "epoch": 25.54552715654952, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 31983 + }, + { + "epoch": 25.54632587859425, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 31984 + }, + { + "epoch": 25.547124600638977, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 31985 + }, + { + "epoch": 25.547923322683705, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 31986 + }, + { + "epoch": 25.548722044728436, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 31987 + }, + { + "epoch": 25.549520766773163, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 31988 + }, + { + "epoch": 25.55031948881789, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 31989 + }, + { + "epoch": 25.55111821086262, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 31990 + }, + { + "epoch": 25.55191693290735, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 31991 + }, + { + "epoch": 25.552715654952078, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 31992 + }, + { + "epoch": 25.553514376996805, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 31993 + }, + { + "epoch": 25.554313099041533, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 31994 + }, + { + "epoch": 25.55511182108626, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 31995 + }, + { + "epoch": 25.55591054313099, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 31996 + }, + { + "epoch": 25.55670926517572, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 31997 + }, + { + "epoch": 25.557507987220447, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 31998 + }, + { + "epoch": 25.558306709265175, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 31999 + }, + { + "epoch": 25.559105431309906, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 32000 + }, + { + "epoch": 25.559904153354633, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32001 + }, + { + "epoch": 25.56070287539936, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 32002 + }, + { + "epoch": 25.56150159744409, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 32003 + }, + { + "epoch": 25.562300319488816, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 32004 + }, + { + "epoch": 25.563099041533548, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 32005 + }, + { + "epoch": 25.563897763578275, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 32006 + }, + { + "epoch": 25.564696485623003, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 32007 + }, + { + "epoch": 25.56549520766773, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 32008 + }, + { + "epoch": 25.56629392971246, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 32009 + }, + { + "epoch": 25.56709265175719, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 32010 + }, + { + "epoch": 25.567891373801917, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0753, + "step": 32011 + }, + { + "epoch": 25.568690095846645, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 32012 + }, + { + "epoch": 25.569488817891372, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 32013 + }, + { + "epoch": 25.570287539936103, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 32014 + }, + { + "epoch": 25.57108626198083, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 32015 + }, + { + "epoch": 25.57188498402556, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 32016 + }, + { + "epoch": 25.572683706070286, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 32017 + }, + { + "epoch": 25.573482428115017, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 32018 + }, + { + "epoch": 25.574281150159745, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 32019 + }, + { + "epoch": 25.575079872204473, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 32020 + }, + { + "epoch": 25.5758785942492, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 32021 + }, + { + "epoch": 25.576677316293928, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 32022 + }, + { + "epoch": 25.57747603833866, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.1083, + "step": 32023 + }, + { + "epoch": 25.578274760383387, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 32024 + }, + { + "epoch": 25.579073482428115, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 32025 + }, + { + "epoch": 25.579872204472842, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 32026 + }, + { + "epoch": 25.580670926517573, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32027 + }, + { + "epoch": 25.5814696485623, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 32028 + }, + { + "epoch": 25.58226837060703, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 32029 + }, + { + "epoch": 25.583067092651756, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 32030 + }, + { + "epoch": 25.583865814696484, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 32031 + }, + { + "epoch": 25.584664536741215, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 32032 + }, + { + "epoch": 25.585463258785943, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 32033 + }, + { + "epoch": 25.58626198083067, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 32034 + }, + { + "epoch": 25.587060702875398, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 32035 + }, + { + "epoch": 25.58785942492013, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 32036 + }, + { + "epoch": 25.588658146964857, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 32037 + }, + { + "epoch": 25.589456869009584, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 32038 + }, + { + "epoch": 25.590255591054312, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 32039 + }, + { + "epoch": 25.591054313099043, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 32040 + }, + { + "epoch": 25.59185303514377, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 32041 + }, + { + "epoch": 25.5926517571885, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32042 + }, + { + "epoch": 25.593450479233226, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 32043 + }, + { + "epoch": 25.594249201277954, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 32044 + }, + { + "epoch": 25.595047923322685, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 32045 + }, + { + "epoch": 25.595846645367413, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32046 + }, + { + "epoch": 25.59664536741214, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 32047 + }, + { + "epoch": 25.597444089456868, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32048 + }, + { + "epoch": 25.5982428115016, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 32049 + }, + { + "epoch": 25.599041533546327, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 32050 + }, + { + "epoch": 25.599840255591054, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 32051 + }, + { + "epoch": 25.600638977635782, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 32052 + }, + { + "epoch": 25.60143769968051, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 32053 + }, + { + "epoch": 25.60223642172524, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 32054 + }, + { + "epoch": 25.60303514376997, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 32055 + }, + { + "epoch": 25.603833865814696, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 32056 + }, + { + "epoch": 25.604632587859424, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 32057 + }, + { + "epoch": 25.605431309904155, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 32058 + }, + { + "epoch": 25.606230031948883, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 32059 + }, + { + "epoch": 25.60702875399361, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32060 + }, + { + "epoch": 25.607827476038338, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 32061 + }, + { + "epoch": 25.608626198083066, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 32062 + }, + { + "epoch": 25.609424920127797, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 32063 + }, + { + "epoch": 25.610223642172524, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 32064 + }, + { + "epoch": 25.611022364217252, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 32065 + }, + { + "epoch": 25.61182108626198, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 32066 + }, + { + "epoch": 25.61261980830671, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 32067 + }, + { + "epoch": 25.61341853035144, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 32068 + }, + { + "epoch": 25.614217252396166, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 32069 + }, + { + "epoch": 25.615015974440894, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 32070 + }, + { + "epoch": 25.61581469648562, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 32071 + }, + { + "epoch": 25.616613418530353, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32072 + }, + { + "epoch": 25.61741214057508, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 32073 + }, + { + "epoch": 25.618210862619808, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 32074 + }, + { + "epoch": 25.619009584664536, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 32075 + }, + { + "epoch": 25.619808306709267, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 32076 + }, + { + "epoch": 25.620607028753994, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0758, + "step": 32077 + }, + { + "epoch": 25.621405750798722, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 32078 + }, + { + "epoch": 25.62220447284345, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 32079 + }, + { + "epoch": 25.623003194888177, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1011, + "step": 32080 + }, + { + "epoch": 25.62380191693291, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 32081 + }, + { + "epoch": 25.624600638977636, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 32082 + }, + { + "epoch": 25.625399361022364, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 32083 + }, + { + "epoch": 25.62619808306709, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 32084 + }, + { + "epoch": 25.626996805111823, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 32085 + }, + { + "epoch": 25.62779552715655, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0753, + "step": 32086 + }, + { + "epoch": 25.628594249201278, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 32087 + }, + { + "epoch": 25.629392971246006, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 32088 + }, + { + "epoch": 25.630191693290733, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 32089 + }, + { + "epoch": 25.630990415335464, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 32090 + }, + { + "epoch": 25.631789137380192, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 32091 + }, + { + "epoch": 25.63258785942492, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 32092 + }, + { + "epoch": 25.633386581469647, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 32093 + }, + { + "epoch": 25.63418530351438, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 32094 + }, + { + "epoch": 25.634984025559106, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32095 + }, + { + "epoch": 25.635782747603834, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 32096 + }, + { + "epoch": 25.63658146964856, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 32097 + }, + { + "epoch": 25.63738019169329, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 32098 + }, + { + "epoch": 25.63817891373802, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 32099 + }, + { + "epoch": 25.638977635782748, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 32100 + }, + { + "epoch": 25.639776357827476, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 32101 + }, + { + "epoch": 25.640575079872203, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 32102 + }, + { + "epoch": 25.641373801916934, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 32103 + }, + { + "epoch": 25.642172523961662, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 32104 + }, + { + "epoch": 25.64297124600639, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 32105 + }, + { + "epoch": 25.643769968051117, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0658, + "step": 32106 + }, + { + "epoch": 25.644568690095845, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 32107 + }, + { + "epoch": 25.645367412140576, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 32108 + }, + { + "epoch": 25.646166134185304, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32109 + }, + { + "epoch": 25.64696485623003, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 32110 + }, + { + "epoch": 25.64776357827476, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 32111 + }, + { + "epoch": 25.64856230031949, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.1032, + "step": 32112 + }, + { + "epoch": 25.649361022364218, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 32113 + }, + { + "epoch": 25.650159744408946, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 32114 + }, + { + "epoch": 25.650958466453673, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 32115 + }, + { + "epoch": 25.6517571884984, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 32116 + }, + { + "epoch": 25.652555910543132, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 32117 + }, + { + "epoch": 25.65335463258786, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 32118 + }, + { + "epoch": 25.654153354632587, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 32119 + }, + { + "epoch": 25.654952076677315, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 32120 + }, + { + "epoch": 25.655750798722046, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 32121 + }, + { + "epoch": 25.656549520766774, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 32122 + }, + { + "epoch": 25.6573482428115, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 32123 + }, + { + "epoch": 25.65814696485623, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 32124 + }, + { + "epoch": 25.658945686900957, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 32125 + }, + { + "epoch": 25.659744408945688, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 32126 + }, + { + "epoch": 25.660543130990416, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 32127 + }, + { + "epoch": 25.661341853035143, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 32128 + }, + { + "epoch": 25.66214057507987, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0703, + "step": 32129 + }, + { + "epoch": 25.662939297124602, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 32130 + }, + { + "epoch": 25.66373801916933, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 32131 + }, + { + "epoch": 25.664536741214057, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 32132 + }, + { + "epoch": 25.665335463258785, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 32133 + }, + { + "epoch": 25.666134185303516, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 32134 + }, + { + "epoch": 25.666932907348244, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 32135 + }, + { + "epoch": 25.66773162939297, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 32136 + }, + { + "epoch": 25.6685303514377, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 32137 + }, + { + "epoch": 25.669329073482427, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 32138 + }, + { + "epoch": 25.670127795527158, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 32139 + }, + { + "epoch": 25.670926517571885, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32140 + }, + { + "epoch": 25.671725239616613, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 32141 + }, + { + "epoch": 25.67252396166134, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 32142 + }, + { + "epoch": 25.673322683706072, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 32143 + }, + { + "epoch": 25.6741214057508, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 32144 + }, + { + "epoch": 25.674920127795527, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 32145 + }, + { + "epoch": 25.675718849840255, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 32146 + }, + { + "epoch": 25.676517571884983, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 32147 + }, + { + "epoch": 25.677316293929714, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 32148 + }, + { + "epoch": 25.67811501597444, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 32149 + }, + { + "epoch": 25.67891373801917, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 32150 + }, + { + "epoch": 25.679712460063897, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 32151 + }, + { + "epoch": 25.680511182108628, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0763, + "step": 32152 + }, + { + "epoch": 25.681309904153355, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 32153 + }, + { + "epoch": 25.682108626198083, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 32154 + }, + { + "epoch": 25.68290734824281, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 32155 + }, + { + "epoch": 25.68370607028754, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 32156 + }, + { + "epoch": 25.68450479233227, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 32157 + }, + { + "epoch": 25.685303514376997, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 32158 + }, + { + "epoch": 25.686102236421725, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 32159 + }, + { + "epoch": 25.686900958466452, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 32160 + }, + { + "epoch": 25.687699680511184, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 32161 + }, + { + "epoch": 25.68849840255591, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 32162 + }, + { + "epoch": 25.68929712460064, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 32163 + }, + { + "epoch": 25.690095846645367, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 32164 + }, + { + "epoch": 25.690894568690094, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 32165 + }, + { + "epoch": 25.691693290734825, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 32166 + }, + { + "epoch": 25.692492012779553, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 32167 + }, + { + "epoch": 25.69329073482428, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 32168 + }, + { + "epoch": 25.69408945686901, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 32169 + }, + { + "epoch": 25.69488817891374, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 32170 + }, + { + "epoch": 25.695686900958467, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 32171 + }, + { + "epoch": 25.696485623003195, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 32172 + }, + { + "epoch": 25.697284345047922, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 32173 + }, + { + "epoch": 25.69808306709265, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 32174 + }, + { + "epoch": 25.69888178913738, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 32175 + }, + { + "epoch": 25.69968051118211, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 32176 + }, + { + "epoch": 25.700479233226837, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 32177 + }, + { + "epoch": 25.701277955271564, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 32178 + }, + { + "epoch": 25.702076677316295, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 32179 + }, + { + "epoch": 25.702875399361023, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 32180 + }, + { + "epoch": 25.70367412140575, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 32181 + }, + { + "epoch": 25.70447284345048, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 32182 + }, + { + "epoch": 25.705271565495206, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 32183 + }, + { + "epoch": 25.706070287539937, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 32184 + }, + { + "epoch": 25.706869009584665, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 32185 + }, + { + "epoch": 25.707667731629392, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 32186 + }, + { + "epoch": 25.70846645367412, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32187 + }, + { + "epoch": 25.70926517571885, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 32188 + }, + { + "epoch": 25.71006389776358, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 32189 + }, + { + "epoch": 25.710862619808307, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 32190 + }, + { + "epoch": 25.711661341853034, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 32191 + }, + { + "epoch": 25.712460063897765, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 32192 + }, + { + "epoch": 25.713258785942493, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 32193 + }, + { + "epoch": 25.71405750798722, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 32194 + }, + { + "epoch": 25.71485623003195, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 32195 + }, + { + "epoch": 25.715654952076676, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 32196 + }, + { + "epoch": 25.716453674121407, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 32197 + }, + { + "epoch": 25.717252396166135, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 32198 + }, + { + "epoch": 25.718051118210862, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 32199 + }, + { + "epoch": 25.71884984025559, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 32200 + }, + { + "epoch": 25.71964856230032, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 32201 + }, + { + "epoch": 25.72044728434505, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 32202 + }, + { + "epoch": 25.721246006389777, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 32203 + }, + { + "epoch": 25.722044728434504, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 32204 + }, + { + "epoch": 25.722843450479232, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 32205 + }, + { + "epoch": 25.723642172523963, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 32206 + }, + { + "epoch": 25.72444089456869, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 32207 + }, + { + "epoch": 25.72523961661342, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 32208 + }, + { + "epoch": 25.726038338658146, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 32209 + }, + { + "epoch": 25.726837060702877, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 32210 + }, + { + "epoch": 25.727635782747605, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 32211 + }, + { + "epoch": 25.728434504792332, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 32212 + }, + { + "epoch": 25.72923322683706, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 32213 + }, + { + "epoch": 25.730031948881788, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32214 + }, + { + "epoch": 25.73083067092652, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 32215 + }, + { + "epoch": 25.731629392971247, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 32216 + }, + { + "epoch": 25.732428115015974, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 32217 + }, + { + "epoch": 25.733226837060702, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 32218 + }, + { + "epoch": 25.734025559105433, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 32219 + }, + { + "epoch": 25.73482428115016, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 32220 + }, + { + "epoch": 25.73562300319489, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 32221 + }, + { + "epoch": 25.736421725239616, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 32222 + }, + { + "epoch": 25.737220447284344, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 32223 + }, + { + "epoch": 25.738019169329075, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 32224 + }, + { + "epoch": 25.738817891373802, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 32225 + }, + { + "epoch": 25.73961661341853, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 32226 + }, + { + "epoch": 25.740415335463258, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 32227 + }, + { + "epoch": 25.74121405750799, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1033, + "step": 32228 + }, + { + "epoch": 25.742012779552716, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 32229 + }, + { + "epoch": 25.742811501597444, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 32230 + }, + { + "epoch": 25.74361022364217, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 32231 + }, + { + "epoch": 25.7444089456869, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 32232 + }, + { + "epoch": 25.74520766773163, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 32233 + }, + { + "epoch": 25.74600638977636, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 32234 + }, + { + "epoch": 25.746805111821086, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 32235 + }, + { + "epoch": 25.747603833865814, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 32236 + }, + { + "epoch": 25.748402555910545, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 32237 + }, + { + "epoch": 25.749201277955272, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 32238 + }, + { + "epoch": 25.75, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 32239 + }, + { + "epoch": 25.750798722044728, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 32240 + }, + { + "epoch": 25.751597444089455, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 32241 + }, + { + "epoch": 25.752396166134186, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 32242 + }, + { + "epoch": 25.753194888178914, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 32243 + }, + { + "epoch": 25.75399361022364, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 32244 + }, + { + "epoch": 25.75479233226837, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 32245 + }, + { + "epoch": 25.7555910543131, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 32246 + }, + { + "epoch": 25.75638977635783, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 32247 + }, + { + "epoch": 25.757188498402556, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 32248 + }, + { + "epoch": 25.757987220447284, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32249 + }, + { + "epoch": 25.75878594249201, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 32250 + }, + { + "epoch": 25.759584664536742, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 32251 + }, + { + "epoch": 25.76038338658147, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 32252 + }, + { + "epoch": 25.761182108626198, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 32253 + }, + { + "epoch": 25.761980830670925, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 32254 + }, + { + "epoch": 25.762779552715656, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 32255 + }, + { + "epoch": 25.763578274760384, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 32256 + }, + { + "epoch": 25.76437699680511, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 32257 + }, + { + "epoch": 25.76517571884984, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 32258 + }, + { + "epoch": 25.765974440894567, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 32259 + }, + { + "epoch": 25.766773162939298, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 32260 + }, + { + "epoch": 25.767571884984026, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 32261 + }, + { + "epoch": 25.768370607028753, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 32262 + }, + { + "epoch": 25.76916932907348, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 32263 + }, + { + "epoch": 25.769968051118212, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 32264 + }, + { + "epoch": 25.77076677316294, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 32265 + }, + { + "epoch": 25.771565495207668, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 32266 + }, + { + "epoch": 25.772364217252395, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 32267 + }, + { + "epoch": 25.773162939297123, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 32268 + }, + { + "epoch": 25.773961661341854, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 32269 + }, + { + "epoch": 25.77476038338658, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 32270 + }, + { + "epoch": 25.77555910543131, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 32271 + }, + { + "epoch": 25.776357827476037, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 32272 + }, + { + "epoch": 25.777156549520768, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 32273 + }, + { + "epoch": 25.777955271565496, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 32274 + }, + { + "epoch": 25.778753993610223, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 32275 + }, + { + "epoch": 25.77955271565495, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 32276 + }, + { + "epoch": 25.78035143769968, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 32277 + }, + { + "epoch": 25.78115015974441, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 32278 + }, + { + "epoch": 25.781948881789138, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 32279 + }, + { + "epoch": 25.782747603833865, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 32280 + }, + { + "epoch": 25.783546325878593, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 32281 + }, + { + "epoch": 25.784345047923324, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 32282 + }, + { + "epoch": 25.78514376996805, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 32283 + }, + { + "epoch": 25.78594249201278, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 32284 + }, + { + "epoch": 25.786741214057507, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 32285 + }, + { + "epoch": 25.787539936102235, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 32286 + }, + { + "epoch": 25.788338658146966, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 32287 + }, + { + "epoch": 25.789137380191693, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 32288 + }, + { + "epoch": 25.78993610223642, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 32289 + }, + { + "epoch": 25.79073482428115, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 32290 + }, + { + "epoch": 25.79153354632588, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 32291 + }, + { + "epoch": 25.792332268370608, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 32292 + }, + { + "epoch": 25.793130990415335, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 32293 + }, + { + "epoch": 25.793929712460063, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 32294 + }, + { + "epoch": 25.794728434504794, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 32295 + }, + { + "epoch": 25.79552715654952, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 32296 + }, + { + "epoch": 25.79632587859425, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 32297 + }, + { + "epoch": 25.797124600638977, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 32298 + }, + { + "epoch": 25.797923322683705, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 32299 + }, + { + "epoch": 25.798722044728436, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32300 + }, + { + "epoch": 25.799520766773163, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 32301 + }, + { + "epoch": 25.80031948881789, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32302 + }, + { + "epoch": 25.80111821086262, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 32303 + }, + { + "epoch": 25.80191693290735, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 32304 + }, + { + "epoch": 25.802715654952078, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 32305 + }, + { + "epoch": 25.803514376996805, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 32306 + }, + { + "epoch": 25.804313099041533, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 32307 + }, + { + "epoch": 25.80511182108626, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 32308 + }, + { + "epoch": 25.80591054313099, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 32309 + }, + { + "epoch": 25.80670926517572, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 32310 + }, + { + "epoch": 25.807507987220447, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 32311 + }, + { + "epoch": 25.808306709265175, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 32312 + }, + { + "epoch": 25.809105431309906, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 32313 + }, + { + "epoch": 25.809904153354633, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 32314 + }, + { + "epoch": 25.81070287539936, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32315 + }, + { + "epoch": 25.81150159744409, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 32316 + }, + { + "epoch": 25.812300319488816, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 32317 + }, + { + "epoch": 25.813099041533548, + "grad_norm": 0.2734375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 32318 + }, + { + "epoch": 25.813897763578275, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 32319 + }, + { + "epoch": 25.814696485623003, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 32320 + }, + { + "epoch": 25.81549520766773, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 32321 + }, + { + "epoch": 25.81629392971246, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 32322 + }, + { + "epoch": 25.81709265175719, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 32323 + }, + { + "epoch": 25.817891373801917, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 32324 + }, + { + "epoch": 25.818690095846645, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 32325 + }, + { + "epoch": 25.819488817891372, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32326 + }, + { + "epoch": 25.820287539936103, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 32327 + }, + { + "epoch": 25.82108626198083, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 32328 + }, + { + "epoch": 25.82188498402556, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32329 + }, + { + "epoch": 25.822683706070286, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 32330 + }, + { + "epoch": 25.823482428115017, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 32331 + }, + { + "epoch": 25.824281150159745, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 32332 + }, + { + "epoch": 25.825079872204473, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 32333 + }, + { + "epoch": 25.8258785942492, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 32334 + }, + { + "epoch": 25.826677316293928, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 32335 + }, + { + "epoch": 25.82747603833866, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 32336 + }, + { + "epoch": 25.828274760383387, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 32337 + }, + { + "epoch": 25.829073482428115, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 32338 + }, + { + "epoch": 25.829872204472842, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 32339 + }, + { + "epoch": 25.830670926517573, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 32340 + }, + { + "epoch": 25.8314696485623, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32341 + }, + { + "epoch": 25.83226837060703, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 32342 + }, + { + "epoch": 25.833067092651756, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32343 + }, + { + "epoch": 25.833865814696484, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32344 + }, + { + "epoch": 25.834664536741215, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 32345 + }, + { + "epoch": 25.835463258785943, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 32346 + }, + { + "epoch": 25.83626198083067, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 32347 + }, + { + "epoch": 25.837060702875398, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 32348 + }, + { + "epoch": 25.83785942492013, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32349 + }, + { + "epoch": 25.838658146964857, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 32350 + }, + { + "epoch": 25.839456869009584, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 32351 + }, + { + "epoch": 25.840255591054312, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 32352 + }, + { + "epoch": 25.841054313099043, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 32353 + }, + { + "epoch": 25.84185303514377, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 32354 + }, + { + "epoch": 25.8426517571885, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0773, + "step": 32355 + }, + { + "epoch": 25.843450479233226, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 32356 + }, + { + "epoch": 25.844249201277954, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 32357 + }, + { + "epoch": 25.845047923322685, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 32358 + }, + { + "epoch": 25.845846645367413, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 32359 + }, + { + "epoch": 25.84664536741214, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 32360 + }, + { + "epoch": 25.847444089456868, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 32361 + }, + { + "epoch": 25.8482428115016, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 32362 + }, + { + "epoch": 25.849041533546327, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 32363 + }, + { + "epoch": 25.849840255591054, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 32364 + }, + { + "epoch": 25.850638977635782, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 32365 + }, + { + "epoch": 25.85143769968051, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 32366 + }, + { + "epoch": 25.85223642172524, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 32367 + }, + { + "epoch": 25.85303514376997, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 32368 + }, + { + "epoch": 25.853833865814696, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32369 + }, + { + "epoch": 25.854632587859424, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 32370 + }, + { + "epoch": 25.855431309904155, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 32371 + }, + { + "epoch": 25.856230031948883, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 32372 + }, + { + "epoch": 25.85702875399361, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 32373 + }, + { + "epoch": 25.857827476038338, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 32374 + }, + { + "epoch": 25.858626198083066, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 32375 + }, + { + "epoch": 25.859424920127797, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 32376 + }, + { + "epoch": 25.860223642172524, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 32377 + }, + { + "epoch": 25.861022364217252, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 32378 + }, + { + "epoch": 25.86182108626198, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 32379 + }, + { + "epoch": 25.86261980830671, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 32380 + }, + { + "epoch": 25.86341853035144, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 32381 + }, + { + "epoch": 25.864217252396166, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 32382 + }, + { + "epoch": 25.865015974440894, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 32383 + }, + { + "epoch": 25.86581469648562, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 32384 + }, + { + "epoch": 25.866613418530353, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32385 + }, + { + "epoch": 25.86741214057508, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 32386 + }, + { + "epoch": 25.868210862619808, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 32387 + }, + { + "epoch": 25.869009584664536, + "grad_norm": 0.287109375, + "learning_rate": 0.0001, + "loss": 1.0743, + "step": 32388 + }, + { + "epoch": 25.869808306709267, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 32389 + }, + { + "epoch": 25.870607028753994, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 32390 + }, + { + "epoch": 25.871405750798722, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 32391 + }, + { + "epoch": 25.87220447284345, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 32392 + }, + { + "epoch": 25.873003194888177, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 32393 + }, + { + "epoch": 25.87380191693291, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 32394 + }, + { + "epoch": 25.874600638977636, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 32395 + }, + { + "epoch": 25.875399361022364, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 32396 + }, + { + "epoch": 25.87619808306709, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 32397 + }, + { + "epoch": 25.876996805111823, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 32398 + }, + { + "epoch": 25.87779552715655, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32399 + }, + { + "epoch": 25.878594249201278, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 32400 + }, + { + "epoch": 25.879392971246006, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 32401 + }, + { + "epoch": 25.880191693290733, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 32402 + }, + { + "epoch": 25.880990415335464, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 32403 + }, + { + "epoch": 25.881789137380192, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 32404 + }, + { + "epoch": 25.88258785942492, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 32405 + }, + { + "epoch": 25.883386581469647, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 32406 + }, + { + "epoch": 25.88418530351438, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 32407 + }, + { + "epoch": 25.884984025559106, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 32408 + }, + { + "epoch": 25.885782747603834, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 32409 + }, + { + "epoch": 25.88658146964856, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 32410 + }, + { + "epoch": 25.88738019169329, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32411 + }, + { + "epoch": 25.88817891373802, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 32412 + }, + { + "epoch": 25.888977635782748, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 32413 + }, + { + "epoch": 25.889776357827476, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 32414 + }, + { + "epoch": 25.890575079872203, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 32415 + }, + { + "epoch": 25.891373801916934, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 32416 + }, + { + "epoch": 25.892172523961662, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 32417 + }, + { + "epoch": 25.89297124600639, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 32418 + }, + { + "epoch": 25.893769968051117, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 32419 + }, + { + "epoch": 25.894568690095845, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 32420 + }, + { + "epoch": 25.895367412140576, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 32421 + }, + { + "epoch": 25.896166134185304, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32422 + }, + { + "epoch": 25.89696485623003, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 32423 + }, + { + "epoch": 25.89776357827476, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 32424 + }, + { + "epoch": 25.89856230031949, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 32425 + }, + { + "epoch": 25.899361022364218, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 32426 + }, + { + "epoch": 25.900159744408946, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 32427 + }, + { + "epoch": 25.900958466453673, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 32428 + }, + { + "epoch": 25.9017571884984, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 32429 + }, + { + "epoch": 25.902555910543132, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 32430 + }, + { + "epoch": 25.90335463258786, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 32431 + }, + { + "epoch": 25.904153354632587, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 32432 + }, + { + "epoch": 25.904952076677315, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 32433 + }, + { + "epoch": 25.905750798722046, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 32434 + }, + { + "epoch": 25.906549520766774, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 32435 + }, + { + "epoch": 25.9073482428115, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 32436 + }, + { + "epoch": 25.90814696485623, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 32437 + }, + { + "epoch": 25.908945686900957, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 32438 + }, + { + "epoch": 25.909744408945688, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 32439 + }, + { + "epoch": 25.910543130990416, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 32440 + }, + { + "epoch": 25.911341853035143, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 32441 + }, + { + "epoch": 25.91214057507987, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 32442 + }, + { + "epoch": 25.912939297124602, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 32443 + }, + { + "epoch": 25.91373801916933, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32444 + }, + { + "epoch": 25.914536741214057, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 32445 + }, + { + "epoch": 25.915335463258785, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 32446 + }, + { + "epoch": 25.916134185303516, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 32447 + }, + { + "epoch": 25.916932907348244, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 32448 + }, + { + "epoch": 25.91773162939297, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 32449 + }, + { + "epoch": 25.9185303514377, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 32450 + }, + { + "epoch": 25.919329073482427, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 32451 + }, + { + "epoch": 25.920127795527158, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 32452 + }, + { + "epoch": 25.920926517571885, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 32453 + }, + { + "epoch": 25.921725239616613, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 32454 + }, + { + "epoch": 25.92252396166134, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 32455 + }, + { + "epoch": 25.923322683706072, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 32456 + }, + { + "epoch": 25.9241214057508, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 32457 + }, + { + "epoch": 25.924920127795527, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 32458 + }, + { + "epoch": 25.925718849840255, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 32459 + }, + { + "epoch": 25.926517571884983, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 32460 + }, + { + "epoch": 25.927316293929714, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 32461 + }, + { + "epoch": 25.92811501597444, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32462 + }, + { + "epoch": 25.92891373801917, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 32463 + }, + { + "epoch": 25.929712460063897, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32464 + }, + { + "epoch": 25.930511182108628, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 32465 + }, + { + "epoch": 25.931309904153355, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 32466 + }, + { + "epoch": 25.932108626198083, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 32467 + }, + { + "epoch": 25.93290734824281, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 32468 + }, + { + "epoch": 25.93370607028754, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 32469 + }, + { + "epoch": 25.93450479233227, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 32470 + }, + { + "epoch": 25.935303514376997, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 32471 + }, + { + "epoch": 25.936102236421725, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 32472 + }, + { + "epoch": 25.936900958466452, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 32473 + }, + { + "epoch": 25.937699680511184, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 32474 + }, + { + "epoch": 25.93849840255591, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 32475 + }, + { + "epoch": 25.93929712460064, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 32476 + }, + { + "epoch": 25.940095846645367, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32477 + }, + { + "epoch": 25.940894568690094, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 32478 + }, + { + "epoch": 25.941693290734825, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 32479 + }, + { + "epoch": 25.942492012779553, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 32480 + }, + { + "epoch": 25.94329073482428, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 32481 + }, + { + "epoch": 25.94408945686901, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 32482 + }, + { + "epoch": 25.94488817891374, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 32483 + }, + { + "epoch": 25.945686900958467, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 32484 + }, + { + "epoch": 25.946485623003195, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 32485 + }, + { + "epoch": 25.947284345047922, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 32486 + }, + { + "epoch": 25.94808306709265, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 32487 + }, + { + "epoch": 25.94888178913738, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32488 + }, + { + "epoch": 25.94968051118211, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 32489 + }, + { + "epoch": 25.950479233226837, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 32490 + }, + { + "epoch": 25.951277955271564, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 32491 + }, + { + "epoch": 25.952076677316295, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 32492 + }, + { + "epoch": 25.952875399361023, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32493 + }, + { + "epoch": 25.95367412140575, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 32494 + }, + { + "epoch": 25.95447284345048, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 32495 + }, + { + "epoch": 25.955271565495206, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 32496 + }, + { + "epoch": 25.956070287539937, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 32497 + }, + { + "epoch": 25.956869009584665, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 32498 + }, + { + "epoch": 25.957667731629392, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 32499 + }, + { + "epoch": 25.95846645367412, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 32500 + }, + { + "epoch": 25.95926517571885, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 32501 + }, + { + "epoch": 25.96006389776358, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 32502 + }, + { + "epoch": 25.960862619808307, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 32503 + }, + { + "epoch": 25.961661341853034, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 32504 + }, + { + "epoch": 25.962460063897765, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 32505 + }, + { + "epoch": 25.963258785942493, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 32506 + }, + { + "epoch": 25.96405750798722, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 32507 + }, + { + "epoch": 25.96485623003195, + "grad_norm": 0.255859375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 32508 + }, + { + "epoch": 25.965654952076676, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 32509 + }, + { + "epoch": 25.966453674121407, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 32510 + }, + { + "epoch": 25.967252396166135, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32511 + }, + { + "epoch": 25.968051118210862, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 32512 + }, + { + "epoch": 25.96884984025559, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 32513 + }, + { + "epoch": 25.96964856230032, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 32514 + }, + { + "epoch": 25.97044728434505, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 32515 + }, + { + "epoch": 25.971246006389777, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 32516 + }, + { + "epoch": 25.972044728434504, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 32517 + }, + { + "epoch": 25.972843450479232, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 32518 + }, + { + "epoch": 25.973642172523963, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 32519 + }, + { + "epoch": 25.97444089456869, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 32520 + }, + { + "epoch": 25.97523961661342, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 32521 + }, + { + "epoch": 25.976038338658146, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 32522 + }, + { + "epoch": 25.976837060702877, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 32523 + }, + { + "epoch": 25.977635782747605, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 32524 + }, + { + "epoch": 25.978434504792332, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 32525 + }, + { + "epoch": 25.97923322683706, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 32526 + }, + { + "epoch": 25.980031948881788, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32527 + }, + { + "epoch": 25.98083067092652, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 32528 + }, + { + "epoch": 25.981629392971247, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 32529 + }, + { + "epoch": 25.982428115015974, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 32530 + }, + { + "epoch": 25.983226837060702, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 32531 + }, + { + "epoch": 25.984025559105433, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 32532 + }, + { + "epoch": 25.98482428115016, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 32533 + }, + { + "epoch": 25.98562300319489, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 32534 + }, + { + "epoch": 25.986421725239616, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 32535 + }, + { + "epoch": 25.987220447284344, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 32536 + }, + { + "epoch": 25.988019169329075, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 32537 + }, + { + "epoch": 25.988817891373802, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 32538 + }, + { + "epoch": 25.98961661341853, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 32539 + }, + { + "epoch": 25.990415335463258, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 32540 + }, + { + "epoch": 25.99121405750799, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 32541 + }, + { + "epoch": 25.992012779552716, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 32542 + }, + { + "epoch": 25.992811501597444, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 32543 + }, + { + "epoch": 25.99361022364217, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 32544 + }, + { + "epoch": 25.9944089456869, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 32545 + }, + { + "epoch": 25.99520766773163, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 32546 + }, + { + "epoch": 25.99600638977636, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0752, + "step": 32547 + }, + { + "epoch": 25.996805111821086, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 32548 + }, + { + "epoch": 25.997603833865814, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 32549 + }, + { + "epoch": 25.998402555910545, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 32550 + }, + { + "epoch": 25.999201277955272, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 32551 + }, + { + "epoch": 26.0, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 32552 + }, + { + "epoch": 26.000798722044728, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 32553 + }, + { + "epoch": 26.001597444089455, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 32554 + }, + { + "epoch": 26.002396166134186, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 32555 + }, + { + "epoch": 26.003194888178914, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 32556 + }, + { + "epoch": 26.00399361022364, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 32557 + }, + { + "epoch": 26.00479233226837, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 32558 + }, + { + "epoch": 26.0055910543131, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 32559 + }, + { + "epoch": 26.00638977635783, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 32560 + }, + { + "epoch": 26.007188498402556, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 32561 + }, + { + "epoch": 26.007987220447284, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 32562 + }, + { + "epoch": 26.00878594249201, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 32563 + }, + { + "epoch": 26.009584664536742, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 32564 + }, + { + "epoch": 26.01038338658147, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 32565 + }, + { + "epoch": 26.011182108626198, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 32566 + }, + { + "epoch": 26.011980830670925, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 32567 + }, + { + "epoch": 26.012779552715656, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 32568 + }, + { + "epoch": 26.013578274760384, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 32569 + }, + { + "epoch": 26.01437699680511, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 32570 + }, + { + "epoch": 26.01517571884984, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 32571 + }, + { + "epoch": 26.015974440894567, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 32572 + }, + { + "epoch": 26.016773162939298, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 32573 + }, + { + "epoch": 26.017571884984026, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 32574 + }, + { + "epoch": 26.018370607028753, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 32575 + }, + { + "epoch": 26.01916932907348, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0742, + "step": 32576 + }, + { + "epoch": 26.019968051118212, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 32577 + }, + { + "epoch": 26.02076677316294, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 32578 + }, + { + "epoch": 26.021565495207668, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 32579 + }, + { + "epoch": 26.022364217252395, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 32580 + }, + { + "epoch": 26.023162939297123, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 32581 + }, + { + "epoch": 26.023961661341854, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 32582 + }, + { + "epoch": 26.02476038338658, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 32583 + }, + { + "epoch": 26.02555910543131, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 32584 + }, + { + "epoch": 26.026357827476037, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 32585 + }, + { + "epoch": 26.027156549520768, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 32586 + }, + { + "epoch": 26.027955271565496, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 32587 + }, + { + "epoch": 26.028753993610223, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 32588 + }, + { + "epoch": 26.02955271565495, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 32589 + }, + { + "epoch": 26.03035143769968, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 32590 + }, + { + "epoch": 26.03115015974441, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 32591 + }, + { + "epoch": 26.031948881789138, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 32592 + }, + { + "epoch": 26.032747603833865, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32593 + }, + { + "epoch": 26.033546325878593, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 32594 + }, + { + "epoch": 26.034345047923324, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 32595 + }, + { + "epoch": 26.03514376996805, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 32596 + }, + { + "epoch": 26.03594249201278, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 32597 + }, + { + "epoch": 26.036741214057507, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 32598 + }, + { + "epoch": 26.037539936102238, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 32599 + }, + { + "epoch": 26.038338658146966, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 32600 + }, + { + "epoch": 26.039137380191693, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 32601 + }, + { + "epoch": 26.03993610223642, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 32602 + }, + { + "epoch": 26.04073482428115, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 32603 + }, + { + "epoch": 26.04153354632588, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 32604 + }, + { + "epoch": 26.042332268370608, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 32605 + }, + { + "epoch": 26.043130990415335, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 32606 + }, + { + "epoch": 26.043929712460063, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 32607 + }, + { + "epoch": 26.044728434504794, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 32608 + }, + { + "epoch": 26.04552715654952, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 32609 + }, + { + "epoch": 26.04632587859425, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 32610 + }, + { + "epoch": 26.047124600638977, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 32611 + }, + { + "epoch": 26.047923322683705, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 32612 + }, + { + "epoch": 26.048722044728436, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 32613 + }, + { + "epoch": 26.049520766773163, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 32614 + }, + { + "epoch": 26.05031948881789, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 32615 + }, + { + "epoch": 26.05111821086262, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 32616 + }, + { + "epoch": 26.05191693290735, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 32617 + }, + { + "epoch": 26.052715654952078, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 32618 + }, + { + "epoch": 26.053514376996805, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 32619 + }, + { + "epoch": 26.054313099041533, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 32620 + }, + { + "epoch": 26.05511182108626, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 32621 + }, + { + "epoch": 26.05591054313099, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 32622 + }, + { + "epoch": 26.05670926517572, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 32623 + }, + { + "epoch": 26.057507987220447, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 32624 + }, + { + "epoch": 26.058306709265175, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 32625 + }, + { + "epoch": 26.059105431309906, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 32626 + }, + { + "epoch": 26.059904153354633, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 32627 + }, + { + "epoch": 26.06070287539936, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 32628 + }, + { + "epoch": 26.06150159744409, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 32629 + }, + { + "epoch": 26.062300319488816, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 32630 + }, + { + "epoch": 26.063099041533548, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 32631 + }, + { + "epoch": 26.063897763578275, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 32632 + }, + { + "epoch": 26.064696485623003, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 32633 + }, + { + "epoch": 26.06549520766773, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 32634 + }, + { + "epoch": 26.06629392971246, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 32635 + }, + { + "epoch": 26.06709265175719, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 32636 + }, + { + "epoch": 26.067891373801917, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 32637 + }, + { + "epoch": 26.068690095846645, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 32638 + }, + { + "epoch": 26.069488817891372, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 32639 + }, + { + "epoch": 26.070287539936103, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 32640 + }, + { + "epoch": 26.07108626198083, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 32641 + }, + { + "epoch": 26.07188498402556, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 32642 + }, + { + "epoch": 26.072683706070286, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 32643 + }, + { + "epoch": 26.073482428115017, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 32644 + }, + { + "epoch": 26.074281150159745, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 32645 + }, + { + "epoch": 26.075079872204473, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 32646 + }, + { + "epoch": 26.0758785942492, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 32647 + }, + { + "epoch": 26.076677316293928, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 32648 + }, + { + "epoch": 26.07747603833866, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 32649 + }, + { + "epoch": 26.078274760383387, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 32650 + }, + { + "epoch": 26.079073482428115, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 32651 + }, + { + "epoch": 26.079872204472842, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 32652 + }, + { + "epoch": 26.080670926517573, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 32653 + }, + { + "epoch": 26.0814696485623, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 32654 + }, + { + "epoch": 26.08226837060703, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32655 + }, + { + "epoch": 26.083067092651756, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 32656 + }, + { + "epoch": 26.083865814696484, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 32657 + }, + { + "epoch": 26.084664536741215, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 32658 + }, + { + "epoch": 26.085463258785943, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 32659 + }, + { + "epoch": 26.08626198083067, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 32660 + }, + { + "epoch": 26.087060702875398, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 32661 + }, + { + "epoch": 26.08785942492013, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 32662 + }, + { + "epoch": 26.088658146964857, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 32663 + }, + { + "epoch": 26.089456869009584, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 32664 + }, + { + "epoch": 26.090255591054312, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32665 + }, + { + "epoch": 26.09105431309904, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 32666 + }, + { + "epoch": 26.09185303514377, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 32667 + }, + { + "epoch": 26.0926517571885, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 32668 + }, + { + "epoch": 26.093450479233226, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 32669 + }, + { + "epoch": 26.094249201277954, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 32670 + }, + { + "epoch": 26.095047923322685, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32671 + }, + { + "epoch": 26.095846645367413, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 32672 + }, + { + "epoch": 26.09664536741214, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32673 + }, + { + "epoch": 26.097444089456868, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 32674 + }, + { + "epoch": 26.0982428115016, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 32675 + }, + { + "epoch": 26.099041533546327, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 32676 + }, + { + "epoch": 26.099840255591054, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 32677 + }, + { + "epoch": 26.100638977635782, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 32678 + }, + { + "epoch": 26.10143769968051, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 32679 + }, + { + "epoch": 26.10223642172524, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 32680 + }, + { + "epoch": 26.10303514376997, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 32681 + }, + { + "epoch": 26.103833865814696, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 32682 + }, + { + "epoch": 26.104632587859424, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 32683 + }, + { + "epoch": 26.105431309904155, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 32684 + }, + { + "epoch": 26.106230031948883, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 32685 + }, + { + "epoch": 26.10702875399361, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 32686 + }, + { + "epoch": 26.107827476038338, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 32687 + }, + { + "epoch": 26.108626198083066, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32688 + }, + { + "epoch": 26.109424920127797, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 32689 + }, + { + "epoch": 26.110223642172524, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 32690 + }, + { + "epoch": 26.111022364217252, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 32691 + }, + { + "epoch": 26.11182108626198, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 32692 + }, + { + "epoch": 26.11261980830671, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 32693 + }, + { + "epoch": 26.11341853035144, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 32694 + }, + { + "epoch": 26.114217252396166, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 32695 + }, + { + "epoch": 26.115015974440894, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 32696 + }, + { + "epoch": 26.11581469648562, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 32697 + }, + { + "epoch": 26.116613418530353, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 32698 + }, + { + "epoch": 26.11741214057508, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 32699 + }, + { + "epoch": 26.118210862619808, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 32700 + }, + { + "epoch": 26.119009584664536, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 32701 + }, + { + "epoch": 26.119808306709267, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 32702 + }, + { + "epoch": 26.120607028753994, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 32703 + }, + { + "epoch": 26.121405750798722, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 32704 + }, + { + "epoch": 26.12220447284345, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 32705 + }, + { + "epoch": 26.123003194888177, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 32706 + }, + { + "epoch": 26.12380191693291, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 32707 + }, + { + "epoch": 26.124600638977636, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 32708 + }, + { + "epoch": 26.125399361022364, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32709 + }, + { + "epoch": 26.12619808306709, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 32710 + }, + { + "epoch": 26.126996805111823, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 32711 + }, + { + "epoch": 26.12779552715655, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 32712 + }, + { + "epoch": 26.128594249201278, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 32713 + }, + { + "epoch": 26.129392971246006, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 32714 + }, + { + "epoch": 26.130191693290733, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 32715 + }, + { + "epoch": 26.130990415335464, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 32716 + }, + { + "epoch": 26.131789137380192, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32717 + }, + { + "epoch": 26.13258785942492, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 32718 + }, + { + "epoch": 26.133386581469647, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 32719 + }, + { + "epoch": 26.13418530351438, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 32720 + }, + { + "epoch": 26.134984025559106, + "grad_norm": 0.30078125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 32721 + }, + { + "epoch": 26.135782747603834, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 32722 + }, + { + "epoch": 26.13658146964856, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 32723 + }, + { + "epoch": 26.13738019169329, + "grad_norm": 0.279296875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 32724 + }, + { + "epoch": 26.13817891373802, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 32725 + }, + { + "epoch": 26.138977635782748, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 32726 + }, + { + "epoch": 26.139776357827476, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 32727 + }, + { + "epoch": 26.140575079872203, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 32728 + }, + { + "epoch": 26.141373801916934, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 32729 + }, + { + "epoch": 26.142172523961662, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 32730 + }, + { + "epoch": 26.14297124600639, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 32731 + }, + { + "epoch": 26.143769968051117, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 32732 + }, + { + "epoch": 26.144568690095845, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 32733 + }, + { + "epoch": 26.145367412140576, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0761, + "step": 32734 + }, + { + "epoch": 26.146166134185304, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.075, + "step": 32735 + }, + { + "epoch": 26.14696485623003, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 32736 + }, + { + "epoch": 26.14776357827476, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 32737 + }, + { + "epoch": 26.14856230031949, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 32738 + }, + { + "epoch": 26.149361022364218, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 32739 + }, + { + "epoch": 26.150159744408946, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 32740 + }, + { + "epoch": 26.150958466453673, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 32741 + }, + { + "epoch": 26.1517571884984, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 32742 + }, + { + "epoch": 26.152555910543132, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 32743 + }, + { + "epoch": 26.15335463258786, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 32744 + }, + { + "epoch": 26.154153354632587, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32745 + }, + { + "epoch": 26.154952076677315, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 32746 + }, + { + "epoch": 26.155750798722046, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 32747 + }, + { + "epoch": 26.156549520766774, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 32748 + }, + { + "epoch": 26.1573482428115, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 32749 + }, + { + "epoch": 26.15814696485623, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 32750 + }, + { + "epoch": 26.15894568690096, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 32751 + }, + { + "epoch": 26.159744408945688, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 32752 + }, + { + "epoch": 26.160543130990416, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 32753 + }, + { + "epoch": 26.161341853035143, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 32754 + }, + { + "epoch": 26.16214057507987, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 32755 + }, + { + "epoch": 26.162939297124602, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32756 + }, + { + "epoch": 26.16373801916933, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 32757 + }, + { + "epoch": 26.164536741214057, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 32758 + }, + { + "epoch": 26.165335463258785, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 32759 + }, + { + "epoch": 26.166134185303516, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 32760 + }, + { + "epoch": 26.166932907348244, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 32761 + }, + { + "epoch": 26.16773162939297, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 32762 + }, + { + "epoch": 26.1685303514377, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 32763 + }, + { + "epoch": 26.169329073482427, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 32764 + }, + { + "epoch": 26.170127795527158, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 32765 + }, + { + "epoch": 26.170926517571885, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 32766 + }, + { + "epoch": 26.171725239616613, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32767 + }, + { + "epoch": 26.17252396166134, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 32768 + }, + { + "epoch": 26.173322683706072, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 32769 + }, + { + "epoch": 26.1741214057508, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 32770 + }, + { + "epoch": 26.174920127795527, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 32771 + }, + { + "epoch": 26.175718849840255, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 32772 + }, + { + "epoch": 26.176517571884983, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 32773 + }, + { + "epoch": 26.177316293929714, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 32774 + }, + { + "epoch": 26.17811501597444, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 32775 + }, + { + "epoch": 26.17891373801917, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 32776 + }, + { + "epoch": 26.179712460063897, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 32777 + }, + { + "epoch": 26.180511182108628, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 32778 + }, + { + "epoch": 26.181309904153355, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 32779 + }, + { + "epoch": 26.182108626198083, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 32780 + }, + { + "epoch": 26.18290734824281, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 32781 + }, + { + "epoch": 26.18370607028754, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 32782 + }, + { + "epoch": 26.18450479233227, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 32783 + }, + { + "epoch": 26.185303514376997, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 32784 + }, + { + "epoch": 26.186102236421725, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 32785 + }, + { + "epoch": 26.186900958466452, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 32786 + }, + { + "epoch": 26.187699680511184, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 32787 + }, + { + "epoch": 26.18849840255591, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 32788 + }, + { + "epoch": 26.18929712460064, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 32789 + }, + { + "epoch": 26.190095846645367, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32790 + }, + { + "epoch": 26.190894568690094, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 32791 + }, + { + "epoch": 26.191693290734825, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 32792 + }, + { + "epoch": 26.192492012779553, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 32793 + }, + { + "epoch": 26.19329073482428, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 32794 + }, + { + "epoch": 26.19408945686901, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32795 + }, + { + "epoch": 26.19488817891374, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 32796 + }, + { + "epoch": 26.195686900958467, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 32797 + }, + { + "epoch": 26.196485623003195, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 32798 + }, + { + "epoch": 26.197284345047922, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 32799 + }, + { + "epoch": 26.19808306709265, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 32800 + }, + { + "epoch": 26.19888178913738, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 32801 + }, + { + "epoch": 26.19968051118211, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 32802 + }, + { + "epoch": 26.200479233226837, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 32803 + }, + { + "epoch": 26.201277955271564, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 32804 + }, + { + "epoch": 26.202076677316295, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 32805 + }, + { + "epoch": 26.202875399361023, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 32806 + }, + { + "epoch": 26.20367412140575, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 32807 + }, + { + "epoch": 26.20447284345048, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 32808 + }, + { + "epoch": 26.205271565495206, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 32809 + }, + { + "epoch": 26.206070287539937, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0773, + "step": 32810 + }, + { + "epoch": 26.206869009584665, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0754, + "step": 32811 + }, + { + "epoch": 26.207667731629392, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 32812 + }, + { + "epoch": 26.20846645367412, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 32813 + }, + { + "epoch": 26.20926517571885, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 32814 + }, + { + "epoch": 26.21006389776358, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 32815 + }, + { + "epoch": 26.210862619808307, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32816 + }, + { + "epoch": 26.211661341853034, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 32817 + }, + { + "epoch": 26.212460063897762, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 32818 + }, + { + "epoch": 26.213258785942493, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 32819 + }, + { + "epoch": 26.21405750798722, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 32820 + }, + { + "epoch": 26.21485623003195, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 32821 + }, + { + "epoch": 26.215654952076676, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 32822 + }, + { + "epoch": 26.216453674121407, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 32823 + }, + { + "epoch": 26.217252396166135, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 32824 + }, + { + "epoch": 26.218051118210862, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 32825 + }, + { + "epoch": 26.21884984025559, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 32826 + }, + { + "epoch": 26.21964856230032, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 32827 + }, + { + "epoch": 26.22044728434505, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 32828 + }, + { + "epoch": 26.221246006389777, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32829 + }, + { + "epoch": 26.222044728434504, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 32830 + }, + { + "epoch": 26.222843450479232, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 32831 + }, + { + "epoch": 26.223642172523963, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 32832 + }, + { + "epoch": 26.22444089456869, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 32833 + }, + { + "epoch": 26.22523961661342, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 32834 + }, + { + "epoch": 26.226038338658146, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 32835 + }, + { + "epoch": 26.226837060702877, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32836 + }, + { + "epoch": 26.227635782747605, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 32837 + }, + { + "epoch": 26.228434504792332, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 32838 + }, + { + "epoch": 26.22923322683706, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 32839 + }, + { + "epoch": 26.230031948881788, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 32840 + }, + { + "epoch": 26.23083067092652, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 32841 + }, + { + "epoch": 26.231629392971247, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 32842 + }, + { + "epoch": 26.232428115015974, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 32843 + }, + { + "epoch": 26.233226837060702, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 32844 + }, + { + "epoch": 26.234025559105433, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 32845 + }, + { + "epoch": 26.23482428115016, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 32846 + }, + { + "epoch": 26.23562300319489, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 32847 + }, + { + "epoch": 26.236421725239616, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 32848 + }, + { + "epoch": 26.237220447284344, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 32849 + }, + { + "epoch": 26.238019169329075, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32850 + }, + { + "epoch": 26.238817891373802, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 32851 + }, + { + "epoch": 26.23961661341853, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 32852 + }, + { + "epoch": 26.240415335463258, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 32853 + }, + { + "epoch": 26.24121405750799, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 32854 + }, + { + "epoch": 26.242012779552716, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 32855 + }, + { + "epoch": 26.242811501597444, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 32856 + }, + { + "epoch": 26.24361022364217, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 32857 + }, + { + "epoch": 26.2444089456869, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 32858 + }, + { + "epoch": 26.24520766773163, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 32859 + }, + { + "epoch": 26.24600638977636, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 32860 + }, + { + "epoch": 26.246805111821086, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 32861 + }, + { + "epoch": 26.247603833865814, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 32862 + }, + { + "epoch": 26.248402555910545, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 32863 + }, + { + "epoch": 26.249201277955272, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 32864 + }, + { + "epoch": 26.25, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 32865 + }, + { + "epoch": 26.250798722044728, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 32866 + }, + { + "epoch": 26.251597444089455, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 32867 + }, + { + "epoch": 26.252396166134186, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 32868 + }, + { + "epoch": 26.253194888178914, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 32869 + }, + { + "epoch": 26.25399361022364, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 32870 + }, + { + "epoch": 26.25479233226837, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 32871 + }, + { + "epoch": 26.2555910543131, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 32872 + }, + { + "epoch": 26.25638977635783, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 32873 + }, + { + "epoch": 26.257188498402556, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32874 + }, + { + "epoch": 26.257987220447284, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 32875 + }, + { + "epoch": 26.25878594249201, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 32876 + }, + { + "epoch": 26.259584664536742, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 32877 + }, + { + "epoch": 26.26038338658147, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 32878 + }, + { + "epoch": 26.261182108626198, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 32879 + }, + { + "epoch": 26.261980830670925, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 32880 + }, + { + "epoch": 26.262779552715656, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 32881 + }, + { + "epoch": 26.263578274760384, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 32882 + }, + { + "epoch": 26.26437699680511, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 32883 + }, + { + "epoch": 26.26517571884984, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 32884 + }, + { + "epoch": 26.265974440894567, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 32885 + }, + { + "epoch": 26.266773162939298, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 32886 + }, + { + "epoch": 26.267571884984026, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 32887 + }, + { + "epoch": 26.268370607028753, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 32888 + }, + { + "epoch": 26.26916932907348, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 32889 + }, + { + "epoch": 26.269968051118212, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 32890 + }, + { + "epoch": 26.27076677316294, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 32891 + }, + { + "epoch": 26.271565495207668, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 32892 + }, + { + "epoch": 26.272364217252395, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 32893 + }, + { + "epoch": 26.273162939297123, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32894 + }, + { + "epoch": 26.273961661341854, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 32895 + }, + { + "epoch": 26.27476038338658, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 32896 + }, + { + "epoch": 26.27555910543131, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 32897 + }, + { + "epoch": 26.276357827476037, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 32898 + }, + { + "epoch": 26.277156549520768, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 32899 + }, + { + "epoch": 26.277955271565496, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 32900 + }, + { + "epoch": 26.278753993610223, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 32901 + }, + { + "epoch": 26.27955271565495, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 32902 + }, + { + "epoch": 26.28035143769968, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 32903 + }, + { + "epoch": 26.28115015974441, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 32904 + }, + { + "epoch": 26.281948881789138, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 32905 + }, + { + "epoch": 26.282747603833865, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 32906 + }, + { + "epoch": 26.283546325878593, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 32907 + }, + { + "epoch": 26.284345047923324, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 32908 + }, + { + "epoch": 26.28514376996805, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 32909 + }, + { + "epoch": 26.28594249201278, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 32910 + }, + { + "epoch": 26.286741214057507, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32911 + }, + { + "epoch": 26.287539936102238, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 32912 + }, + { + "epoch": 26.288338658146966, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 32913 + }, + { + "epoch": 26.289137380191693, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 32914 + }, + { + "epoch": 26.28993610223642, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 32915 + }, + { + "epoch": 26.29073482428115, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 32916 + }, + { + "epoch": 26.29153354632588, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 32917 + }, + { + "epoch": 26.292332268370608, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 32918 + }, + { + "epoch": 26.293130990415335, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 32919 + }, + { + "epoch": 26.293929712460063, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 32920 + }, + { + "epoch": 26.294728434504794, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 32921 + }, + { + "epoch": 26.29552715654952, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 32922 + }, + { + "epoch": 26.29632587859425, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 32923 + }, + { + "epoch": 26.297124600638977, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 32924 + }, + { + "epoch": 26.297923322683705, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 32925 + }, + { + "epoch": 26.298722044728436, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 32926 + }, + { + "epoch": 26.299520766773163, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 32927 + }, + { + "epoch": 26.30031948881789, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 32928 + }, + { + "epoch": 26.30111821086262, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 32929 + }, + { + "epoch": 26.30191693290735, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 32930 + }, + { + "epoch": 26.302715654952078, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 32931 + }, + { + "epoch": 26.303514376996805, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 32932 + }, + { + "epoch": 26.304313099041533, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 32933 + }, + { + "epoch": 26.30511182108626, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 32934 + }, + { + "epoch": 26.30591054313099, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 32935 + }, + { + "epoch": 26.30670926517572, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 32936 + }, + { + "epoch": 26.307507987220447, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 32937 + }, + { + "epoch": 26.308306709265175, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 32938 + }, + { + "epoch": 26.309105431309906, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 32939 + }, + { + "epoch": 26.309904153354633, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 32940 + }, + { + "epoch": 26.31070287539936, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 32941 + }, + { + "epoch": 26.31150159744409, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 32942 + }, + { + "epoch": 26.312300319488816, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 32943 + }, + { + "epoch": 26.313099041533548, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 32944 + }, + { + "epoch": 26.313897763578275, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 32945 + }, + { + "epoch": 26.314696485623003, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 32946 + }, + { + "epoch": 26.31549520766773, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 32947 + }, + { + "epoch": 26.31629392971246, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 32948 + }, + { + "epoch": 26.31709265175719, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 32949 + }, + { + "epoch": 26.317891373801917, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 32950 + }, + { + "epoch": 26.318690095846645, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 32951 + }, + { + "epoch": 26.319488817891372, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 32952 + }, + { + "epoch": 26.320287539936103, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 32953 + }, + { + "epoch": 26.32108626198083, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 32954 + }, + { + "epoch": 26.32188498402556, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 32955 + }, + { + "epoch": 26.322683706070286, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.1056, + "step": 32956 + }, + { + "epoch": 26.323482428115017, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 32957 + }, + { + "epoch": 26.324281150159745, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 32958 + }, + { + "epoch": 26.325079872204473, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 32959 + }, + { + "epoch": 26.3258785942492, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 32960 + }, + { + "epoch": 26.326677316293928, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 32961 + }, + { + "epoch": 26.32747603833866, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 32962 + }, + { + "epoch": 26.328274760383387, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 32963 + }, + { + "epoch": 26.329073482428115, + "grad_norm": 0.09375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 32964 + }, + { + "epoch": 26.329872204472842, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 32965 + }, + { + "epoch": 26.330670926517573, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 32966 + }, + { + "epoch": 26.3314696485623, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 32967 + }, + { + "epoch": 26.33226837060703, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 32968 + }, + { + "epoch": 26.333067092651756, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 32969 + }, + { + "epoch": 26.333865814696484, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 32970 + }, + { + "epoch": 26.334664536741215, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 32971 + }, + { + "epoch": 26.335463258785943, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 32972 + }, + { + "epoch": 26.33626198083067, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 32973 + }, + { + "epoch": 26.337060702875398, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 32974 + }, + { + "epoch": 26.33785942492013, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 32975 + }, + { + "epoch": 26.338658146964857, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 32976 + }, + { + "epoch": 26.339456869009584, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 32977 + }, + { + "epoch": 26.340255591054312, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 32978 + }, + { + "epoch": 26.341054313099043, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 32979 + }, + { + "epoch": 26.34185303514377, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 32980 + }, + { + "epoch": 26.3426517571885, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 32981 + }, + { + "epoch": 26.343450479233226, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 32982 + }, + { + "epoch": 26.344249201277954, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 32983 + }, + { + "epoch": 26.345047923322685, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 32984 + }, + { + "epoch": 26.345846645367413, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 32985 + }, + { + "epoch": 26.34664536741214, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 32986 + }, + { + "epoch": 26.347444089456868, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 32987 + }, + { + "epoch": 26.3482428115016, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 32988 + }, + { + "epoch": 26.349041533546327, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 32989 + }, + { + "epoch": 26.349840255591054, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 32990 + }, + { + "epoch": 26.350638977635782, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 32991 + }, + { + "epoch": 26.35143769968051, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 32992 + }, + { + "epoch": 26.35223642172524, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 32993 + }, + { + "epoch": 26.35303514376997, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 32994 + }, + { + "epoch": 26.353833865814696, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 32995 + }, + { + "epoch": 26.354632587859424, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 32996 + }, + { + "epoch": 26.355431309904155, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 32997 + }, + { + "epoch": 26.356230031948883, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 32998 + }, + { + "epoch": 26.35702875399361, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 32999 + }, + { + "epoch": 26.357827476038338, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 33000 + }, + { + "epoch": 26.358626198083066, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 33001 + }, + { + "epoch": 26.359424920127797, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 33002 + }, + { + "epoch": 26.360223642172524, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 33003 + }, + { + "epoch": 26.361022364217252, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 33004 + }, + { + "epoch": 26.36182108626198, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 33005 + }, + { + "epoch": 26.36261980830671, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 33006 + }, + { + "epoch": 26.36341853035144, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33007 + }, + { + "epoch": 26.364217252396166, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 33008 + }, + { + "epoch": 26.365015974440894, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 33009 + }, + { + "epoch": 26.36581469648562, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 33010 + }, + { + "epoch": 26.366613418530353, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 33011 + }, + { + "epoch": 26.36741214057508, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 33012 + }, + { + "epoch": 26.368210862619808, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 33013 + }, + { + "epoch": 26.369009584664536, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 33014 + }, + { + "epoch": 26.369808306709267, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 33015 + }, + { + "epoch": 26.370607028753994, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 33016 + }, + { + "epoch": 26.371405750798722, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 33017 + }, + { + "epoch": 26.37220447284345, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 33018 + }, + { + "epoch": 26.373003194888177, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 33019 + }, + { + "epoch": 26.37380191693291, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 33020 + }, + { + "epoch": 26.374600638977636, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 33021 + }, + { + "epoch": 26.375399361022364, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 33022 + }, + { + "epoch": 26.37619808306709, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 33023 + }, + { + "epoch": 26.376996805111823, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 33024 + }, + { + "epoch": 26.37779552715655, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 33025 + }, + { + "epoch": 26.378594249201278, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33026 + }, + { + "epoch": 26.379392971246006, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 33027 + }, + { + "epoch": 26.380191693290733, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 33028 + }, + { + "epoch": 26.380990415335464, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 33029 + }, + { + "epoch": 26.381789137380192, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 33030 + }, + { + "epoch": 26.38258785942492, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 33031 + }, + { + "epoch": 26.383386581469647, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 33032 + }, + { + "epoch": 26.38418530351438, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 33033 + }, + { + "epoch": 26.384984025559106, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 33034 + }, + { + "epoch": 26.385782747603834, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 33035 + }, + { + "epoch": 26.38658146964856, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 33036 + }, + { + "epoch": 26.38738019169329, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 33037 + }, + { + "epoch": 26.38817891373802, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 33038 + }, + { + "epoch": 26.388977635782748, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0765, + "step": 33039 + }, + { + "epoch": 26.389776357827476, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.1016, + "step": 33040 + }, + { + "epoch": 26.390575079872203, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 33041 + }, + { + "epoch": 26.391373801916934, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 33042 + }, + { + "epoch": 26.392172523961662, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 33043 + }, + { + "epoch": 26.39297124600639, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 33044 + }, + { + "epoch": 26.393769968051117, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 33045 + }, + { + "epoch": 26.394568690095845, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 33046 + }, + { + "epoch": 26.395367412140576, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33047 + }, + { + "epoch": 26.396166134185304, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0739, + "step": 33048 + }, + { + "epoch": 26.39696485623003, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 33049 + }, + { + "epoch": 26.39776357827476, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33050 + }, + { + "epoch": 26.39856230031949, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 33051 + }, + { + "epoch": 26.399361022364218, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 33052 + }, + { + "epoch": 26.400159744408946, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 33053 + }, + { + "epoch": 26.400958466453673, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 33054 + }, + { + "epoch": 26.4017571884984, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 33055 + }, + { + "epoch": 26.402555910543132, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 33056 + }, + { + "epoch": 26.40335463258786, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33057 + }, + { + "epoch": 26.404153354632587, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 33058 + }, + { + "epoch": 26.404952076677315, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 33059 + }, + { + "epoch": 26.405750798722046, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 33060 + }, + { + "epoch": 26.406549520766774, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 33061 + }, + { + "epoch": 26.4073482428115, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 33062 + }, + { + "epoch": 26.40814696485623, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 33063 + }, + { + "epoch": 26.408945686900957, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 33064 + }, + { + "epoch": 26.409744408945688, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 33065 + }, + { + "epoch": 26.410543130990416, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 33066 + }, + { + "epoch": 26.411341853035143, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 33067 + }, + { + "epoch": 26.41214057507987, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 33068 + }, + { + "epoch": 26.412939297124602, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 33069 + }, + { + "epoch": 26.41373801916933, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 33070 + }, + { + "epoch": 26.414536741214057, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33071 + }, + { + "epoch": 26.415335463258785, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 33072 + }, + { + "epoch": 26.416134185303516, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 33073 + }, + { + "epoch": 26.416932907348244, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 33074 + }, + { + "epoch": 26.41773162939297, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 33075 + }, + { + "epoch": 26.4185303514377, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 33076 + }, + { + "epoch": 26.419329073482427, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33077 + }, + { + "epoch": 26.420127795527158, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 33078 + }, + { + "epoch": 26.420926517571885, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 33079 + }, + { + "epoch": 26.421725239616613, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 33080 + }, + { + "epoch": 26.42252396166134, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 33081 + }, + { + "epoch": 26.423322683706072, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 33082 + }, + { + "epoch": 26.4241214057508, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 33083 + }, + { + "epoch": 26.424920127795527, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 33084 + }, + { + "epoch": 26.425718849840255, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 33085 + }, + { + "epoch": 26.426517571884983, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 33086 + }, + { + "epoch": 26.427316293929714, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 33087 + }, + { + "epoch": 26.42811501597444, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 33088 + }, + { + "epoch": 26.42891373801917, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 33089 + }, + { + "epoch": 26.429712460063897, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 33090 + }, + { + "epoch": 26.430511182108628, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33091 + }, + { + "epoch": 26.431309904153355, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 33092 + }, + { + "epoch": 26.432108626198083, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 33093 + }, + { + "epoch": 26.43290734824281, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 33094 + }, + { + "epoch": 26.43370607028754, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 33095 + }, + { + "epoch": 26.43450479233227, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 33096 + }, + { + "epoch": 26.435303514376997, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 33097 + }, + { + "epoch": 26.436102236421725, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 33098 + }, + { + "epoch": 26.436900958466452, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 33099 + }, + { + "epoch": 26.437699680511184, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 33100 + }, + { + "epoch": 26.43849840255591, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 33101 + }, + { + "epoch": 26.43929712460064, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 33102 + }, + { + "epoch": 26.440095846645367, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 33103 + }, + { + "epoch": 26.440894568690094, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 33104 + }, + { + "epoch": 26.441693290734825, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 33105 + }, + { + "epoch": 26.442492012779553, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 33106 + }, + { + "epoch": 26.44329073482428, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 33107 + }, + { + "epoch": 26.44408945686901, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 33108 + }, + { + "epoch": 26.44488817891374, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 33109 + }, + { + "epoch": 26.445686900958467, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 33110 + }, + { + "epoch": 26.446485623003195, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 33111 + }, + { + "epoch": 26.447284345047922, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 33112 + }, + { + "epoch": 26.44808306709265, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 33113 + }, + { + "epoch": 26.44888178913738, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 33114 + }, + { + "epoch": 26.44968051118211, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 33115 + }, + { + "epoch": 26.450479233226837, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 33116 + }, + { + "epoch": 26.451277955271564, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33117 + }, + { + "epoch": 26.452076677316295, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33118 + }, + { + "epoch": 26.452875399361023, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0722, + "step": 33119 + }, + { + "epoch": 26.45367412140575, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 33120 + }, + { + "epoch": 26.45447284345048, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 33121 + }, + { + "epoch": 26.455271565495206, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 33122 + }, + { + "epoch": 26.456070287539937, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 33123 + }, + { + "epoch": 26.456869009584665, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 33124 + }, + { + "epoch": 26.457667731629392, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 33125 + }, + { + "epoch": 26.45846645367412, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 33126 + }, + { + "epoch": 26.45926517571885, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 33127 + }, + { + "epoch": 26.46006389776358, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 33128 + }, + { + "epoch": 26.460862619808307, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 33129 + }, + { + "epoch": 26.461661341853034, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 33130 + }, + { + "epoch": 26.462460063897762, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 33131 + }, + { + "epoch": 26.463258785942493, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 33132 + }, + { + "epoch": 26.46405750798722, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 33133 + }, + { + "epoch": 26.46485623003195, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 33134 + }, + { + "epoch": 26.465654952076676, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 33135 + }, + { + "epoch": 26.466453674121407, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 33136 + }, + { + "epoch": 26.467252396166135, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33137 + }, + { + "epoch": 26.468051118210862, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 33138 + }, + { + "epoch": 26.46884984025559, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 33139 + }, + { + "epoch": 26.46964856230032, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 33140 + }, + { + "epoch": 26.47044728434505, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 33141 + }, + { + "epoch": 26.471246006389777, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 33142 + }, + { + "epoch": 26.472044728434504, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 33143 + }, + { + "epoch": 26.472843450479232, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 33144 + }, + { + "epoch": 26.473642172523963, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 33145 + }, + { + "epoch": 26.47444089456869, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 33146 + }, + { + "epoch": 26.47523961661342, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 33147 + }, + { + "epoch": 26.476038338658146, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 33148 + }, + { + "epoch": 26.476837060702877, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 33149 + }, + { + "epoch": 26.477635782747605, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 33150 + }, + { + "epoch": 26.478434504792332, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 33151 + }, + { + "epoch": 26.47923322683706, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 33152 + }, + { + "epoch": 26.480031948881788, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 33153 + }, + { + "epoch": 26.48083067092652, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 33154 + }, + { + "epoch": 26.481629392971247, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 33155 + }, + { + "epoch": 26.482428115015974, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 33156 + }, + { + "epoch": 26.483226837060702, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33157 + }, + { + "epoch": 26.484025559105433, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 33158 + }, + { + "epoch": 26.48482428115016, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33159 + }, + { + "epoch": 26.48562300319489, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33160 + }, + { + "epoch": 26.486421725239616, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 33161 + }, + { + "epoch": 26.487220447284344, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 33162 + }, + { + "epoch": 26.488019169329075, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 33163 + }, + { + "epoch": 26.488817891373802, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 33164 + }, + { + "epoch": 26.48961661341853, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 33165 + }, + { + "epoch": 26.490415335463258, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 33166 + }, + { + "epoch": 26.49121405750799, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 33167 + }, + { + "epoch": 26.492012779552716, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 33168 + }, + { + "epoch": 26.492811501597444, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 33169 + }, + { + "epoch": 26.49361022364217, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 33170 + }, + { + "epoch": 26.4944089456869, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33171 + }, + { + "epoch": 26.49520766773163, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 33172 + }, + { + "epoch": 26.49600638977636, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 33173 + }, + { + "epoch": 26.496805111821086, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 33174 + }, + { + "epoch": 26.497603833865814, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 33175 + }, + { + "epoch": 26.498402555910545, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 33176 + }, + { + "epoch": 26.499201277955272, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33177 + }, + { + "epoch": 26.5, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 33178 + }, + { + "epoch": 26.500798722044728, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 33179 + }, + { + "epoch": 26.501597444089455, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 33180 + }, + { + "epoch": 26.502396166134186, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 33181 + }, + { + "epoch": 26.503194888178914, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 33182 + }, + { + "epoch": 26.50399361022364, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 33183 + }, + { + "epoch": 26.50479233226837, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 33184 + }, + { + "epoch": 26.5055910543131, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 33185 + }, + { + "epoch": 26.50638977635783, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 33186 + }, + { + "epoch": 26.507188498402556, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 33187 + }, + { + "epoch": 26.507987220447284, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 33188 + }, + { + "epoch": 26.50878594249201, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 33189 + }, + { + "epoch": 26.509584664536742, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 33190 + }, + { + "epoch": 26.51038338658147, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 33191 + }, + { + "epoch": 26.511182108626198, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 33192 + }, + { + "epoch": 26.511980830670925, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 33193 + }, + { + "epoch": 26.512779552715656, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 33194 + }, + { + "epoch": 26.513578274760384, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 33195 + }, + { + "epoch": 26.51437699680511, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 33196 + }, + { + "epoch": 26.51517571884984, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 33197 + }, + { + "epoch": 26.515974440894567, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 33198 + }, + { + "epoch": 26.516773162939298, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 33199 + }, + { + "epoch": 26.517571884984026, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 33200 + }, + { + "epoch": 26.518370607028753, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 33201 + }, + { + "epoch": 26.51916932907348, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33202 + }, + { + "epoch": 26.519968051118212, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 33203 + }, + { + "epoch": 26.52076677316294, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 33204 + }, + { + "epoch": 26.521565495207668, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 33205 + }, + { + "epoch": 26.522364217252395, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 33206 + }, + { + "epoch": 26.523162939297123, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 33207 + }, + { + "epoch": 26.523961661341854, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 33208 + }, + { + "epoch": 26.52476038338658, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 33209 + }, + { + "epoch": 26.52555910543131, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 33210 + }, + { + "epoch": 26.526357827476037, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 33211 + }, + { + "epoch": 26.527156549520768, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 33212 + }, + { + "epoch": 26.527955271565496, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 33213 + }, + { + "epoch": 26.528753993610223, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 33214 + }, + { + "epoch": 26.52955271565495, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 33215 + }, + { + "epoch": 26.53035143769968, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 33216 + }, + { + "epoch": 26.53115015974441, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 33217 + }, + { + "epoch": 26.531948881789138, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 33218 + }, + { + "epoch": 26.532747603833865, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 33219 + }, + { + "epoch": 26.533546325878593, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33220 + }, + { + "epoch": 26.534345047923324, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 33221 + }, + { + "epoch": 26.53514376996805, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 33222 + }, + { + "epoch": 26.53594249201278, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 33223 + }, + { + "epoch": 26.536741214057507, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 33224 + }, + { + "epoch": 26.537539936102235, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 33225 + }, + { + "epoch": 26.538338658146966, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33226 + }, + { + "epoch": 26.539137380191693, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 33227 + }, + { + "epoch": 26.53993610223642, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 33228 + }, + { + "epoch": 26.54073482428115, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 33229 + }, + { + "epoch": 26.54153354632588, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 33230 + }, + { + "epoch": 26.542332268370608, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 33231 + }, + { + "epoch": 26.543130990415335, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 33232 + }, + { + "epoch": 26.543929712460063, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 33233 + }, + { + "epoch": 26.544728434504794, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 33234 + }, + { + "epoch": 26.54552715654952, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 33235 + }, + { + "epoch": 26.54632587859425, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 33236 + }, + { + "epoch": 26.547124600638977, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 33237 + }, + { + "epoch": 26.547923322683705, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 33238 + }, + { + "epoch": 26.548722044728436, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 33239 + }, + { + "epoch": 26.549520766773163, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 33240 + }, + { + "epoch": 26.55031948881789, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 33241 + }, + { + "epoch": 26.55111821086262, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 33242 + }, + { + "epoch": 26.55191693290735, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 33243 + }, + { + "epoch": 26.552715654952078, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 33244 + }, + { + "epoch": 26.553514376996805, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33245 + }, + { + "epoch": 26.554313099041533, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 33246 + }, + { + "epoch": 26.55511182108626, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33247 + }, + { + "epoch": 26.55591054313099, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 33248 + }, + { + "epoch": 26.55670926517572, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 33249 + }, + { + "epoch": 26.557507987220447, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33250 + }, + { + "epoch": 26.558306709265175, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 33251 + }, + { + "epoch": 26.559105431309906, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 33252 + }, + { + "epoch": 26.559904153354633, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 33253 + }, + { + "epoch": 26.56070287539936, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 33254 + }, + { + "epoch": 26.56150159744409, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 33255 + }, + { + "epoch": 26.562300319488816, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 33256 + }, + { + "epoch": 26.563099041533548, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33257 + }, + { + "epoch": 26.563897763578275, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 33258 + }, + { + "epoch": 26.564696485623003, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 33259 + }, + { + "epoch": 26.56549520766773, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 33260 + }, + { + "epoch": 26.56629392971246, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 33261 + }, + { + "epoch": 26.56709265175719, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 33262 + }, + { + "epoch": 26.567891373801917, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 33263 + }, + { + "epoch": 26.568690095846645, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 33264 + }, + { + "epoch": 26.569488817891372, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 33265 + }, + { + "epoch": 26.570287539936103, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 33266 + }, + { + "epoch": 26.57108626198083, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 33267 + }, + { + "epoch": 26.57188498402556, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 33268 + }, + { + "epoch": 26.572683706070286, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 33269 + }, + { + "epoch": 26.573482428115017, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33270 + }, + { + "epoch": 26.574281150159745, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 33271 + }, + { + "epoch": 26.575079872204473, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 33272 + }, + { + "epoch": 26.5758785942492, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33273 + }, + { + "epoch": 26.576677316293928, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 33274 + }, + { + "epoch": 26.57747603833866, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 33275 + }, + { + "epoch": 26.578274760383387, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 33276 + }, + { + "epoch": 26.579073482428115, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 33277 + }, + { + "epoch": 26.579872204472842, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 33278 + }, + { + "epoch": 26.580670926517573, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 33279 + }, + { + "epoch": 26.5814696485623, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 33280 + }, + { + "epoch": 26.58226837060703, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 33281 + }, + { + "epoch": 26.583067092651756, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 33282 + }, + { + "epoch": 26.583865814696484, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 33283 + }, + { + "epoch": 26.584664536741215, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 33284 + }, + { + "epoch": 26.585463258785943, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 33285 + }, + { + "epoch": 26.58626198083067, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 33286 + }, + { + "epoch": 26.587060702875398, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 33287 + }, + { + "epoch": 26.58785942492013, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 33288 + }, + { + "epoch": 26.588658146964857, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 33289 + }, + { + "epoch": 26.589456869009584, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 33290 + }, + { + "epoch": 26.590255591054312, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 33291 + }, + { + "epoch": 26.591054313099043, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 33292 + }, + { + "epoch": 26.59185303514377, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 33293 + }, + { + "epoch": 26.5926517571885, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 33294 + }, + { + "epoch": 26.593450479233226, + "grad_norm": 0.09521484375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 33295 + }, + { + "epoch": 26.594249201277954, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 33296 + }, + { + "epoch": 26.595047923322685, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33297 + }, + { + "epoch": 26.595846645367413, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 33298 + }, + { + "epoch": 26.59664536741214, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 33299 + }, + { + "epoch": 26.597444089456868, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 33300 + }, + { + "epoch": 26.5982428115016, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 33301 + }, + { + "epoch": 26.599041533546327, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 33302 + }, + { + "epoch": 26.599840255591054, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33303 + }, + { + "epoch": 26.600638977635782, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 33304 + }, + { + "epoch": 26.60143769968051, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 33305 + }, + { + "epoch": 26.60223642172524, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 33306 + }, + { + "epoch": 26.60303514376997, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 33307 + }, + { + "epoch": 26.603833865814696, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 33308 + }, + { + "epoch": 26.604632587859424, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 33309 + }, + { + "epoch": 26.605431309904155, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 33310 + }, + { + "epoch": 26.606230031948883, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 33311 + }, + { + "epoch": 26.60702875399361, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 33312 + }, + { + "epoch": 26.607827476038338, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 33313 + }, + { + "epoch": 26.608626198083066, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 33314 + }, + { + "epoch": 26.609424920127797, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 33315 + }, + { + "epoch": 26.610223642172524, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33316 + }, + { + "epoch": 26.611022364217252, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 33317 + }, + { + "epoch": 26.61182108626198, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0994, + "step": 33318 + }, + { + "epoch": 26.61261980830671, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.1009, + "step": 33319 + }, + { + "epoch": 26.61341853035144, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 33320 + }, + { + "epoch": 26.614217252396166, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 33321 + }, + { + "epoch": 26.615015974440894, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 33322 + }, + { + "epoch": 26.61581469648562, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.1035, + "step": 33323 + }, + { + "epoch": 26.616613418530353, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 33324 + }, + { + "epoch": 26.61741214057508, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 33325 + }, + { + "epoch": 26.618210862619808, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 33326 + }, + { + "epoch": 26.619009584664536, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 33327 + }, + { + "epoch": 26.619808306709267, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 33328 + }, + { + "epoch": 26.620607028753994, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 33329 + }, + { + "epoch": 26.621405750798722, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 33330 + }, + { + "epoch": 26.62220447284345, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 33331 + }, + { + "epoch": 26.623003194888177, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 33332 + }, + { + "epoch": 26.62380191693291, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 33333 + }, + { + "epoch": 26.624600638977636, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 33334 + }, + { + "epoch": 26.625399361022364, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 33335 + }, + { + "epoch": 26.62619808306709, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 33336 + }, + { + "epoch": 26.626996805111823, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 33337 + }, + { + "epoch": 26.62779552715655, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 33338 + }, + { + "epoch": 26.628594249201278, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 33339 + }, + { + "epoch": 26.629392971246006, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 33340 + }, + { + "epoch": 26.630191693290733, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0742, + "step": 33341 + }, + { + "epoch": 26.630990415335464, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 33342 + }, + { + "epoch": 26.631789137380192, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 33343 + }, + { + "epoch": 26.63258785942492, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 33344 + }, + { + "epoch": 26.633386581469647, + "grad_norm": 0.271484375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 33345 + }, + { + "epoch": 26.63418530351438, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33346 + }, + { + "epoch": 26.634984025559106, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 33347 + }, + { + "epoch": 26.635782747603834, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 33348 + }, + { + "epoch": 26.63658146964856, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.076, + "step": 33349 + }, + { + "epoch": 26.63738019169329, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33350 + }, + { + "epoch": 26.63817891373802, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 33351 + }, + { + "epoch": 26.638977635782748, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 33352 + }, + { + "epoch": 26.639776357827476, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 33353 + }, + { + "epoch": 26.640575079872203, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 33354 + }, + { + "epoch": 26.641373801916934, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 33355 + }, + { + "epoch": 26.642172523961662, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 33356 + }, + { + "epoch": 26.64297124600639, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 33357 + }, + { + "epoch": 26.643769968051117, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 33358 + }, + { + "epoch": 26.644568690095845, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 33359 + }, + { + "epoch": 26.645367412140576, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 33360 + }, + { + "epoch": 26.646166134185304, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 33361 + }, + { + "epoch": 26.64696485623003, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 33362 + }, + { + "epoch": 26.64776357827476, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 33363 + }, + { + "epoch": 26.64856230031949, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 33364 + }, + { + "epoch": 26.649361022364218, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 33365 + }, + { + "epoch": 26.650159744408946, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 33366 + }, + { + "epoch": 26.650958466453673, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 33367 + }, + { + "epoch": 26.6517571884984, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 33368 + }, + { + "epoch": 26.652555910543132, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 33369 + }, + { + "epoch": 26.65335463258786, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 33370 + }, + { + "epoch": 26.654153354632587, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 33371 + }, + { + "epoch": 26.654952076677315, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 33372 + }, + { + "epoch": 26.655750798722046, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 33373 + }, + { + "epoch": 26.656549520766774, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 33374 + }, + { + "epoch": 26.6573482428115, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 33375 + }, + { + "epoch": 26.65814696485623, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 33376 + }, + { + "epoch": 26.658945686900957, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 33377 + }, + { + "epoch": 26.659744408945688, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 33378 + }, + { + "epoch": 26.660543130990416, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 33379 + }, + { + "epoch": 26.661341853035143, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 33380 + }, + { + "epoch": 26.66214057507987, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 33381 + }, + { + "epoch": 26.662939297124602, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 33382 + }, + { + "epoch": 26.66373801916933, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 33383 + }, + { + "epoch": 26.664536741214057, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 33384 + }, + { + "epoch": 26.665335463258785, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 33385 + }, + { + "epoch": 26.666134185303516, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 33386 + }, + { + "epoch": 26.666932907348244, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 33387 + }, + { + "epoch": 26.66773162939297, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 33388 + }, + { + "epoch": 26.6685303514377, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 33389 + }, + { + "epoch": 26.669329073482427, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 33390 + }, + { + "epoch": 26.670127795527158, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 33391 + }, + { + "epoch": 26.670926517571885, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 33392 + }, + { + "epoch": 26.671725239616613, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 33393 + }, + { + "epoch": 26.67252396166134, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 33394 + }, + { + "epoch": 26.673322683706072, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 33395 + }, + { + "epoch": 26.6741214057508, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 33396 + }, + { + "epoch": 26.674920127795527, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 33397 + }, + { + "epoch": 26.675718849840255, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 33398 + }, + { + "epoch": 26.676517571884983, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 33399 + }, + { + "epoch": 26.677316293929714, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 33400 + }, + { + "epoch": 26.67811501597444, + "grad_norm": 0.310546875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 33401 + }, + { + "epoch": 26.67891373801917, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 33402 + }, + { + "epoch": 26.679712460063897, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 33403 + }, + { + "epoch": 26.680511182108628, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 33404 + }, + { + "epoch": 26.681309904153355, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 33405 + }, + { + "epoch": 26.682108626198083, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 33406 + }, + { + "epoch": 26.68290734824281, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 33407 + }, + { + "epoch": 26.68370607028754, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 33408 + }, + { + "epoch": 26.68450479233227, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 33409 + }, + { + "epoch": 26.685303514376997, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 33410 + }, + { + "epoch": 26.686102236421725, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 33411 + }, + { + "epoch": 26.686900958466452, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 33412 + }, + { + "epoch": 26.687699680511184, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 33413 + }, + { + "epoch": 26.68849840255591, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 33414 + }, + { + "epoch": 26.68929712460064, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 33415 + }, + { + "epoch": 26.690095846645367, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33416 + }, + { + "epoch": 26.690894568690094, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 33417 + }, + { + "epoch": 26.691693290734825, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 33418 + }, + { + "epoch": 26.692492012779553, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 33419 + }, + { + "epoch": 26.69329073482428, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 33420 + }, + { + "epoch": 26.69408945686901, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 33421 + }, + { + "epoch": 26.69488817891374, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 33422 + }, + { + "epoch": 26.695686900958467, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 33423 + }, + { + "epoch": 26.696485623003195, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33424 + }, + { + "epoch": 26.697284345047922, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 33425 + }, + { + "epoch": 26.69808306709265, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 33426 + }, + { + "epoch": 26.69888178913738, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 33427 + }, + { + "epoch": 26.69968051118211, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 33428 + }, + { + "epoch": 26.700479233226837, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 33429 + }, + { + "epoch": 26.701277955271564, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 33430 + }, + { + "epoch": 26.702076677316295, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 33431 + }, + { + "epoch": 26.702875399361023, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 33432 + }, + { + "epoch": 26.70367412140575, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 33433 + }, + { + "epoch": 26.70447284345048, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 33434 + }, + { + "epoch": 26.705271565495206, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 33435 + }, + { + "epoch": 26.706070287539937, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 33436 + }, + { + "epoch": 26.706869009584665, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 33437 + }, + { + "epoch": 26.707667731629392, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 33438 + }, + { + "epoch": 26.70846645367412, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 33439 + }, + { + "epoch": 26.70926517571885, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 33440 + }, + { + "epoch": 26.71006389776358, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 33441 + }, + { + "epoch": 26.710862619808307, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 33442 + }, + { + "epoch": 26.711661341853034, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0725, + "step": 33443 + }, + { + "epoch": 26.712460063897765, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 33444 + }, + { + "epoch": 26.713258785942493, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 33445 + }, + { + "epoch": 26.71405750798722, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 33446 + }, + { + "epoch": 26.71485623003195, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 33447 + }, + { + "epoch": 26.715654952076676, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 33448 + }, + { + "epoch": 26.716453674121407, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 33449 + }, + { + "epoch": 26.717252396166135, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 33450 + }, + { + "epoch": 26.718051118210862, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 33451 + }, + { + "epoch": 26.71884984025559, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 33452 + }, + { + "epoch": 26.71964856230032, + "grad_norm": 0.09814453125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 33453 + }, + { + "epoch": 26.72044728434505, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 33454 + }, + { + "epoch": 26.721246006389777, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33455 + }, + { + "epoch": 26.722044728434504, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 33456 + }, + { + "epoch": 26.722843450479232, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 33457 + }, + { + "epoch": 26.723642172523963, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 33458 + }, + { + "epoch": 26.72444089456869, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 33459 + }, + { + "epoch": 26.72523961661342, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 33460 + }, + { + "epoch": 26.726038338658146, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 33461 + }, + { + "epoch": 26.726837060702877, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 33462 + }, + { + "epoch": 26.727635782747605, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.072, + "step": 33463 + }, + { + "epoch": 26.728434504792332, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 33464 + }, + { + "epoch": 26.72923322683706, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 33465 + }, + { + "epoch": 26.730031948881788, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 33466 + }, + { + "epoch": 26.73083067092652, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 33467 + }, + { + "epoch": 26.731629392971247, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 33468 + }, + { + "epoch": 26.732428115015974, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 33469 + }, + { + "epoch": 26.733226837060702, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 33470 + }, + { + "epoch": 26.734025559105433, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33471 + }, + { + "epoch": 26.73482428115016, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 33472 + }, + { + "epoch": 26.73562300319489, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 33473 + }, + { + "epoch": 26.736421725239616, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 33474 + }, + { + "epoch": 26.737220447284344, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 33475 + }, + { + "epoch": 26.738019169329075, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 33476 + }, + { + "epoch": 26.738817891373802, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 33477 + }, + { + "epoch": 26.73961661341853, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33478 + }, + { + "epoch": 26.740415335463258, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 33479 + }, + { + "epoch": 26.74121405750799, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 33480 + }, + { + "epoch": 26.742012779552716, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 33481 + }, + { + "epoch": 26.742811501597444, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 33482 + }, + { + "epoch": 26.74361022364217, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 33483 + }, + { + "epoch": 26.7444089456869, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 33484 + }, + { + "epoch": 26.74520766773163, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 33485 + }, + { + "epoch": 26.74600638977636, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 33486 + }, + { + "epoch": 26.746805111821086, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 33487 + }, + { + "epoch": 26.747603833865814, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 33488 + }, + { + "epoch": 26.748402555910545, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 33489 + }, + { + "epoch": 26.749201277955272, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 33490 + }, + { + "epoch": 26.75, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 33491 + }, + { + "epoch": 26.750798722044728, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 33492 + }, + { + "epoch": 26.751597444089455, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 33493 + }, + { + "epoch": 26.752396166134186, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 33494 + }, + { + "epoch": 26.753194888178914, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 33495 + }, + { + "epoch": 26.75399361022364, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 33496 + }, + { + "epoch": 26.75479233226837, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 33497 + }, + { + "epoch": 26.7555910543131, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 33498 + }, + { + "epoch": 26.75638977635783, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 33499 + }, + { + "epoch": 26.757188498402556, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 33500 + }, + { + "epoch": 26.757987220447284, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0756, + "step": 33501 + }, + { + "epoch": 26.75878594249201, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 33502 + }, + { + "epoch": 26.759584664536742, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 33503 + }, + { + "epoch": 26.76038338658147, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 33504 + }, + { + "epoch": 26.761182108626198, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 33505 + }, + { + "epoch": 26.761980830670925, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 33506 + }, + { + "epoch": 26.762779552715656, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 33507 + }, + { + "epoch": 26.763578274760384, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 33508 + }, + { + "epoch": 26.76437699680511, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 33509 + }, + { + "epoch": 26.76517571884984, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 33510 + }, + { + "epoch": 26.765974440894567, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 33511 + }, + { + "epoch": 26.766773162939298, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 33512 + }, + { + "epoch": 26.767571884984026, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 33513 + }, + { + "epoch": 26.768370607028753, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33514 + }, + { + "epoch": 26.76916932907348, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 33515 + }, + { + "epoch": 26.769968051118212, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33516 + }, + { + "epoch": 26.77076677316294, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 33517 + }, + { + "epoch": 26.771565495207668, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 33518 + }, + { + "epoch": 26.772364217252395, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 33519 + }, + { + "epoch": 26.773162939297123, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 33520 + }, + { + "epoch": 26.773961661341854, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 33521 + }, + { + "epoch": 26.77476038338658, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 33522 + }, + { + "epoch": 26.77555910543131, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 33523 + }, + { + "epoch": 26.776357827476037, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 33524 + }, + { + "epoch": 26.777156549520768, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 33525 + }, + { + "epoch": 26.777955271565496, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 33526 + }, + { + "epoch": 26.778753993610223, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 33527 + }, + { + "epoch": 26.77955271565495, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33528 + }, + { + "epoch": 26.78035143769968, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 33529 + }, + { + "epoch": 26.78115015974441, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 33530 + }, + { + "epoch": 26.781948881789138, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 33531 + }, + { + "epoch": 26.782747603833865, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 33532 + }, + { + "epoch": 26.783546325878593, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 33533 + }, + { + "epoch": 26.784345047923324, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 33534 + }, + { + "epoch": 26.78514376996805, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 33535 + }, + { + "epoch": 26.78594249201278, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 33536 + }, + { + "epoch": 26.786741214057507, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 33537 + }, + { + "epoch": 26.787539936102235, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 33538 + }, + { + "epoch": 26.788338658146966, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 33539 + }, + { + "epoch": 26.789137380191693, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33540 + }, + { + "epoch": 26.78993610223642, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 33541 + }, + { + "epoch": 26.79073482428115, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 33542 + }, + { + "epoch": 26.79153354632588, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 33543 + }, + { + "epoch": 26.792332268370608, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 33544 + }, + { + "epoch": 26.793130990415335, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 33545 + }, + { + "epoch": 26.793929712460063, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33546 + }, + { + "epoch": 26.794728434504794, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 33547 + }, + { + "epoch": 26.79552715654952, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 33548 + }, + { + "epoch": 26.79632587859425, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 33549 + }, + { + "epoch": 26.797124600638977, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 33550 + }, + { + "epoch": 26.797923322683705, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 33551 + }, + { + "epoch": 26.798722044728436, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 33552 + }, + { + "epoch": 26.799520766773163, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 33553 + }, + { + "epoch": 26.80031948881789, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 33554 + }, + { + "epoch": 26.80111821086262, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 33555 + }, + { + "epoch": 26.80191693290735, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 33556 + }, + { + "epoch": 26.802715654952078, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 33557 + }, + { + "epoch": 26.803514376996805, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 33558 + }, + { + "epoch": 26.804313099041533, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 33559 + }, + { + "epoch": 26.80511182108626, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 33560 + }, + { + "epoch": 26.80591054313099, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 33561 + }, + { + "epoch": 26.80670926517572, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 33562 + }, + { + "epoch": 26.807507987220447, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33563 + }, + { + "epoch": 26.808306709265175, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 33564 + }, + { + "epoch": 26.809105431309906, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33565 + }, + { + "epoch": 26.809904153354633, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 33566 + }, + { + "epoch": 26.81070287539936, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 33567 + }, + { + "epoch": 26.81150159744409, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 33568 + }, + { + "epoch": 26.812300319488816, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 33569 + }, + { + "epoch": 26.813099041533548, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 33570 + }, + { + "epoch": 26.813897763578275, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33571 + }, + { + "epoch": 26.814696485623003, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 33572 + }, + { + "epoch": 26.81549520766773, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 33573 + }, + { + "epoch": 26.81629392971246, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 33574 + }, + { + "epoch": 26.81709265175719, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 33575 + }, + { + "epoch": 26.817891373801917, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 33576 + }, + { + "epoch": 26.818690095846645, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 33577 + }, + { + "epoch": 26.819488817891372, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 33578 + }, + { + "epoch": 26.820287539936103, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 33579 + }, + { + "epoch": 26.82108626198083, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 33580 + }, + { + "epoch": 26.82188498402556, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 33581 + }, + { + "epoch": 26.822683706070286, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 33582 + }, + { + "epoch": 26.823482428115017, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 33583 + }, + { + "epoch": 26.824281150159745, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33584 + }, + { + "epoch": 26.825079872204473, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 33585 + }, + { + "epoch": 26.8258785942492, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 33586 + }, + { + "epoch": 26.826677316293928, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 33587 + }, + { + "epoch": 26.82747603833866, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 33588 + }, + { + "epoch": 26.828274760383387, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 33589 + }, + { + "epoch": 26.829073482428115, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 33590 + }, + { + "epoch": 26.829872204472842, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 33591 + }, + { + "epoch": 26.830670926517573, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 33592 + }, + { + "epoch": 26.8314696485623, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 33593 + }, + { + "epoch": 26.83226837060703, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 33594 + }, + { + "epoch": 26.833067092651756, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 33595 + }, + { + "epoch": 26.833865814696484, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 33596 + }, + { + "epoch": 26.834664536741215, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 33597 + }, + { + "epoch": 26.835463258785943, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 33598 + }, + { + "epoch": 26.83626198083067, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 33599 + }, + { + "epoch": 26.837060702875398, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 33600 + }, + { + "epoch": 26.83785942492013, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33601 + }, + { + "epoch": 26.838658146964857, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 33602 + }, + { + "epoch": 26.839456869009584, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 33603 + }, + { + "epoch": 26.840255591054312, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33604 + }, + { + "epoch": 26.841054313099043, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 33605 + }, + { + "epoch": 26.84185303514377, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 33606 + }, + { + "epoch": 26.8426517571885, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 33607 + }, + { + "epoch": 26.843450479233226, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 33608 + }, + { + "epoch": 26.844249201277954, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 33609 + }, + { + "epoch": 26.845047923322685, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 33610 + }, + { + "epoch": 26.845846645367413, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 33611 + }, + { + "epoch": 26.84664536741214, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 33612 + }, + { + "epoch": 26.847444089456868, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 33613 + }, + { + "epoch": 26.8482428115016, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33614 + }, + { + "epoch": 26.849041533546327, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 33615 + }, + { + "epoch": 26.849840255591054, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 33616 + }, + { + "epoch": 26.850638977635782, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 33617 + }, + { + "epoch": 26.85143769968051, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 33618 + }, + { + "epoch": 26.85223642172524, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 33619 + }, + { + "epoch": 26.85303514376997, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 33620 + }, + { + "epoch": 26.853833865814696, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 33621 + }, + { + "epoch": 26.854632587859424, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 33622 + }, + { + "epoch": 26.855431309904155, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 33623 + }, + { + "epoch": 26.856230031948883, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 33624 + }, + { + "epoch": 26.85702875399361, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 33625 + }, + { + "epoch": 26.857827476038338, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 33626 + }, + { + "epoch": 26.858626198083066, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 33627 + }, + { + "epoch": 26.859424920127797, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 33628 + }, + { + "epoch": 26.860223642172524, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 33629 + }, + { + "epoch": 26.861022364217252, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 33630 + }, + { + "epoch": 26.86182108626198, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 33631 + }, + { + "epoch": 26.86261980830671, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 33632 + }, + { + "epoch": 26.86341853035144, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 33633 + }, + { + "epoch": 26.864217252396166, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 33634 + }, + { + "epoch": 26.865015974440894, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 33635 + }, + { + "epoch": 26.86581469648562, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 33636 + }, + { + "epoch": 26.866613418530353, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 33637 + }, + { + "epoch": 26.86741214057508, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 33638 + }, + { + "epoch": 26.868210862619808, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 33639 + }, + { + "epoch": 26.869009584664536, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 33640 + }, + { + "epoch": 26.869808306709267, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 33641 + }, + { + "epoch": 26.870607028753994, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 33642 + }, + { + "epoch": 26.871405750798722, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33643 + }, + { + "epoch": 26.87220447284345, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 33644 + }, + { + "epoch": 26.873003194888177, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 33645 + }, + { + "epoch": 26.87380191693291, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 33646 + }, + { + "epoch": 26.874600638977636, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 33647 + }, + { + "epoch": 26.875399361022364, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 33648 + }, + { + "epoch": 26.87619808306709, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 33649 + }, + { + "epoch": 26.876996805111823, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 33650 + }, + { + "epoch": 26.87779552715655, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 33651 + }, + { + "epoch": 26.878594249201278, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 33652 + }, + { + "epoch": 26.879392971246006, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0773, + "step": 33653 + }, + { + "epoch": 26.880191693290733, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 33654 + }, + { + "epoch": 26.880990415335464, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 33655 + }, + { + "epoch": 26.881789137380192, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 33656 + }, + { + "epoch": 26.88258785942492, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33657 + }, + { + "epoch": 26.883386581469647, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 33658 + }, + { + "epoch": 26.88418530351438, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 33659 + }, + { + "epoch": 26.884984025559106, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 33660 + }, + { + "epoch": 26.885782747603834, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 33661 + }, + { + "epoch": 26.88658146964856, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 33662 + }, + { + "epoch": 26.88738019169329, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 33663 + }, + { + "epoch": 26.88817891373802, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33664 + }, + { + "epoch": 26.888977635782748, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33665 + }, + { + "epoch": 26.889776357827476, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 33666 + }, + { + "epoch": 26.890575079872203, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 33667 + }, + { + "epoch": 26.891373801916934, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 33668 + }, + { + "epoch": 26.892172523961662, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 33669 + }, + { + "epoch": 26.89297124600639, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 33670 + }, + { + "epoch": 26.893769968051117, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 33671 + }, + { + "epoch": 26.894568690095845, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 33672 + }, + { + "epoch": 26.895367412140576, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 33673 + }, + { + "epoch": 26.896166134185304, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 33674 + }, + { + "epoch": 26.89696485623003, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 33675 + }, + { + "epoch": 26.89776357827476, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 33676 + }, + { + "epoch": 26.89856230031949, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 33677 + }, + { + "epoch": 26.899361022364218, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 33678 + }, + { + "epoch": 26.900159744408946, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 33679 + }, + { + "epoch": 26.900958466453673, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 33680 + }, + { + "epoch": 26.9017571884984, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 33681 + }, + { + "epoch": 26.902555910543132, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 33682 + }, + { + "epoch": 26.90335463258786, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 33683 + }, + { + "epoch": 26.904153354632587, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 33684 + }, + { + "epoch": 26.904952076677315, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 33685 + }, + { + "epoch": 26.905750798722046, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 33686 + }, + { + "epoch": 26.906549520766774, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 33687 + }, + { + "epoch": 26.9073482428115, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 33688 + }, + { + "epoch": 26.90814696485623, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 33689 + }, + { + "epoch": 26.908945686900957, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 33690 + }, + { + "epoch": 26.909744408945688, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 33691 + }, + { + "epoch": 26.910543130990416, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 33692 + }, + { + "epoch": 26.911341853035143, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 33693 + }, + { + "epoch": 26.91214057507987, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 33694 + }, + { + "epoch": 26.912939297124602, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 33695 + }, + { + "epoch": 26.91373801916933, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 33696 + }, + { + "epoch": 26.914536741214057, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 33697 + }, + { + "epoch": 26.915335463258785, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 33698 + }, + { + "epoch": 26.916134185303516, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 33699 + }, + { + "epoch": 26.916932907348244, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 33700 + }, + { + "epoch": 26.91773162939297, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 33701 + }, + { + "epoch": 26.9185303514377, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 33702 + }, + { + "epoch": 26.919329073482427, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33703 + }, + { + "epoch": 26.920127795527158, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 33704 + }, + { + "epoch": 26.920926517571885, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 33705 + }, + { + "epoch": 26.921725239616613, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 33706 + }, + { + "epoch": 26.92252396166134, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 33707 + }, + { + "epoch": 26.923322683706072, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 33708 + }, + { + "epoch": 26.9241214057508, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 33709 + }, + { + "epoch": 26.924920127795527, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33710 + }, + { + "epoch": 26.925718849840255, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0732, + "step": 33711 + }, + { + "epoch": 26.926517571884983, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 33712 + }, + { + "epoch": 26.927316293929714, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 33713 + }, + { + "epoch": 26.92811501597444, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 33714 + }, + { + "epoch": 26.92891373801917, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 33715 + }, + { + "epoch": 26.929712460063897, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 33716 + }, + { + "epoch": 26.930511182108628, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 33717 + }, + { + "epoch": 26.931309904153355, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 33718 + }, + { + "epoch": 26.932108626198083, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 33719 + }, + { + "epoch": 26.93290734824281, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 33720 + }, + { + "epoch": 26.93370607028754, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 33721 + }, + { + "epoch": 26.93450479233227, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0762, + "step": 33722 + }, + { + "epoch": 26.935303514376997, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 33723 + }, + { + "epoch": 26.936102236421725, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 33724 + }, + { + "epoch": 26.936900958466452, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 33725 + }, + { + "epoch": 26.937699680511184, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 33726 + }, + { + "epoch": 26.93849840255591, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 33727 + }, + { + "epoch": 26.93929712460064, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 33728 + }, + { + "epoch": 26.940095846645367, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 33729 + }, + { + "epoch": 26.940894568690094, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 33730 + }, + { + "epoch": 26.941693290734825, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 33731 + }, + { + "epoch": 26.942492012779553, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 33732 + }, + { + "epoch": 26.94329073482428, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 33733 + }, + { + "epoch": 26.94408945686901, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33734 + }, + { + "epoch": 26.94488817891374, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 33735 + }, + { + "epoch": 26.945686900958467, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 33736 + }, + { + "epoch": 26.946485623003195, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33737 + }, + { + "epoch": 26.947284345047922, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 33738 + }, + { + "epoch": 26.94808306709265, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 33739 + }, + { + "epoch": 26.94888178913738, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 33740 + }, + { + "epoch": 26.94968051118211, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 33741 + }, + { + "epoch": 26.950479233226837, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 33742 + }, + { + "epoch": 26.951277955271564, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 33743 + }, + { + "epoch": 26.952076677316295, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 33744 + }, + { + "epoch": 26.952875399361023, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 33745 + }, + { + "epoch": 26.95367412140575, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 33746 + }, + { + "epoch": 26.95447284345048, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33747 + }, + { + "epoch": 26.955271565495206, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 33748 + }, + { + "epoch": 26.956070287539937, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 33749 + }, + { + "epoch": 26.956869009584665, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 33750 + }, + { + "epoch": 26.957667731629392, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 33751 + }, + { + "epoch": 26.95846645367412, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 33752 + }, + { + "epoch": 26.95926517571885, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 33753 + }, + { + "epoch": 26.96006389776358, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33754 + }, + { + "epoch": 26.960862619808307, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 33755 + }, + { + "epoch": 26.961661341853034, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 33756 + }, + { + "epoch": 26.962460063897765, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 33757 + }, + { + "epoch": 26.963258785942493, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 33758 + }, + { + "epoch": 26.96405750798722, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 33759 + }, + { + "epoch": 26.96485623003195, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 33760 + }, + { + "epoch": 26.965654952076676, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 33761 + }, + { + "epoch": 26.966453674121407, + "grad_norm": 0.09228515625, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 33762 + }, + { + "epoch": 26.967252396166135, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 33763 + }, + { + "epoch": 26.968051118210862, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 33764 + }, + { + "epoch": 26.96884984025559, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 33765 + }, + { + "epoch": 26.96964856230032, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 33766 + }, + { + "epoch": 26.97044728434505, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 33767 + }, + { + "epoch": 26.971246006389777, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 33768 + }, + { + "epoch": 26.972044728434504, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0767, + "step": 33769 + }, + { + "epoch": 26.972843450479232, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 33770 + }, + { + "epoch": 26.973642172523963, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 33771 + }, + { + "epoch": 26.97444089456869, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 33772 + }, + { + "epoch": 26.97523961661342, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 33773 + }, + { + "epoch": 26.976038338658146, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 33774 + }, + { + "epoch": 26.976837060702877, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 33775 + }, + { + "epoch": 26.977635782747605, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 33776 + }, + { + "epoch": 26.978434504792332, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0753, + "step": 33777 + }, + { + "epoch": 26.97923322683706, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 33778 + }, + { + "epoch": 26.980031948881788, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 33779 + }, + { + "epoch": 26.98083067092652, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 33780 + }, + { + "epoch": 26.981629392971247, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 33781 + }, + { + "epoch": 26.982428115015974, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 33782 + }, + { + "epoch": 26.983226837060702, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 33783 + }, + { + "epoch": 26.984025559105433, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 33784 + }, + { + "epoch": 26.98482428115016, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 33785 + }, + { + "epoch": 26.98562300319489, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 33786 + }, + { + "epoch": 26.986421725239616, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 33787 + }, + { + "epoch": 26.987220447284344, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 33788 + }, + { + "epoch": 26.988019169329075, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 33789 + }, + { + "epoch": 26.988817891373802, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 33790 + }, + { + "epoch": 26.98961661341853, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 33791 + }, + { + "epoch": 26.990415335463258, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 33792 + }, + { + "epoch": 26.99121405750799, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 33793 + }, + { + "epoch": 26.992012779552716, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 33794 + }, + { + "epoch": 26.992811501597444, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 33795 + }, + { + "epoch": 26.99361022364217, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 33796 + }, + { + "epoch": 26.9944089456869, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 33797 + }, + { + "epoch": 26.99520766773163, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 33798 + }, + { + "epoch": 26.99600638977636, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 33799 + }, + { + "epoch": 26.996805111821086, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 33800 + }, + { + "epoch": 26.997603833865814, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 33801 + }, + { + "epoch": 26.998402555910545, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 33802 + }, + { + "epoch": 26.999201277955272, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 33803 + }, + { + "epoch": 27.0, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 33804 + }, + { + "epoch": 27.000798722044728, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.0741, + "step": 33805 + }, + { + "epoch": 27.001597444089455, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 33806 + }, + { + "epoch": 27.002396166134186, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 33807 + }, + { + "epoch": 27.003194888178914, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 33808 + }, + { + "epoch": 27.00399361022364, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0743, + "step": 33809 + }, + { + "epoch": 27.00479233226837, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 33810 + }, + { + "epoch": 27.0055910543131, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 33811 + }, + { + "epoch": 27.00638977635783, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 33812 + }, + { + "epoch": 27.007188498402556, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 33813 + }, + { + "epoch": 27.007987220447284, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 33814 + }, + { + "epoch": 27.00878594249201, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 33815 + }, + { + "epoch": 27.009584664536742, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 33816 + }, + { + "epoch": 27.01038338658147, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 33817 + }, + { + "epoch": 27.011182108626198, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 33818 + }, + { + "epoch": 27.011980830670925, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 33819 + }, + { + "epoch": 27.012779552715656, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33820 + }, + { + "epoch": 27.013578274760384, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 33821 + }, + { + "epoch": 27.01437699680511, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 33822 + }, + { + "epoch": 27.01517571884984, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 33823 + }, + { + "epoch": 27.015974440894567, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 33824 + }, + { + "epoch": 27.016773162939298, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 33825 + }, + { + "epoch": 27.017571884984026, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 33826 + }, + { + "epoch": 27.018370607028753, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 33827 + }, + { + "epoch": 27.01916932907348, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 33828 + }, + { + "epoch": 27.019968051118212, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 33829 + }, + { + "epoch": 27.02076677316294, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 33830 + }, + { + "epoch": 27.021565495207668, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 33831 + }, + { + "epoch": 27.022364217252395, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 33832 + }, + { + "epoch": 27.023162939297123, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 33833 + }, + { + "epoch": 27.023961661341854, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 33834 + }, + { + "epoch": 27.02476038338658, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 33835 + }, + { + "epoch": 27.02555910543131, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 33836 + }, + { + "epoch": 27.026357827476037, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 33837 + }, + { + "epoch": 27.027156549520768, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 33838 + }, + { + "epoch": 27.027955271565496, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 33839 + }, + { + "epoch": 27.028753993610223, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 33840 + }, + { + "epoch": 27.02955271565495, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 33841 + }, + { + "epoch": 27.03035143769968, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 33842 + }, + { + "epoch": 27.03115015974441, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 33843 + }, + { + "epoch": 27.031948881789138, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0751, + "step": 33844 + }, + { + "epoch": 27.032747603833865, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 33845 + }, + { + "epoch": 27.033546325878593, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 33846 + }, + { + "epoch": 27.034345047923324, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33847 + }, + { + "epoch": 27.03514376996805, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 33848 + }, + { + "epoch": 27.03594249201278, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 33849 + }, + { + "epoch": 27.036741214057507, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33850 + }, + { + "epoch": 27.037539936102238, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33851 + }, + { + "epoch": 27.038338658146966, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 33852 + }, + { + "epoch": 27.039137380191693, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 33853 + }, + { + "epoch": 27.03993610223642, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 33854 + }, + { + "epoch": 27.04073482428115, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 33855 + }, + { + "epoch": 27.04153354632588, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33856 + }, + { + "epoch": 27.042332268370608, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 33857 + }, + { + "epoch": 27.043130990415335, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 33858 + }, + { + "epoch": 27.043929712460063, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 33859 + }, + { + "epoch": 27.044728434504794, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 33860 + }, + { + "epoch": 27.04552715654952, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33861 + }, + { + "epoch": 27.04632587859425, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 33862 + }, + { + "epoch": 27.047124600638977, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 33863 + }, + { + "epoch": 27.047923322683705, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 33864 + }, + { + "epoch": 27.048722044728436, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 33865 + }, + { + "epoch": 27.049520766773163, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33866 + }, + { + "epoch": 27.05031948881789, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 33867 + }, + { + "epoch": 27.05111821086262, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 33868 + }, + { + "epoch": 27.05191693290735, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 33869 + }, + { + "epoch": 27.052715654952078, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 33870 + }, + { + "epoch": 27.053514376996805, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 33871 + }, + { + "epoch": 27.054313099041533, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 33872 + }, + { + "epoch": 27.05511182108626, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 33873 + }, + { + "epoch": 27.05591054313099, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 33874 + }, + { + "epoch": 27.05670926517572, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 33875 + }, + { + "epoch": 27.057507987220447, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33876 + }, + { + "epoch": 27.058306709265175, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 33877 + }, + { + "epoch": 27.059105431309906, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 33878 + }, + { + "epoch": 27.059904153354633, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 33879 + }, + { + "epoch": 27.06070287539936, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 33880 + }, + { + "epoch": 27.06150159744409, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 33881 + }, + { + "epoch": 27.062300319488816, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 33882 + }, + { + "epoch": 27.063099041533548, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 33883 + }, + { + "epoch": 27.063897763578275, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 33884 + }, + { + "epoch": 27.064696485623003, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 33885 + }, + { + "epoch": 27.06549520766773, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 33886 + }, + { + "epoch": 27.06629392971246, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 33887 + }, + { + "epoch": 27.06709265175719, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 33888 + }, + { + "epoch": 27.067891373801917, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 33889 + }, + { + "epoch": 27.068690095846645, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33890 + }, + { + "epoch": 27.069488817891372, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 33891 + }, + { + "epoch": 27.070287539936103, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 33892 + }, + { + "epoch": 27.07108626198083, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 33893 + }, + { + "epoch": 27.07188498402556, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 33894 + }, + { + "epoch": 27.072683706070286, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 33895 + }, + { + "epoch": 27.073482428115017, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 33896 + }, + { + "epoch": 27.074281150159745, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 33897 + }, + { + "epoch": 27.075079872204473, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 33898 + }, + { + "epoch": 27.0758785942492, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 33899 + }, + { + "epoch": 27.076677316293928, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 33900 + }, + { + "epoch": 27.07747603833866, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 33901 + }, + { + "epoch": 27.078274760383387, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 33902 + }, + { + "epoch": 27.079073482428115, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 33903 + }, + { + "epoch": 27.079872204472842, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 33904 + }, + { + "epoch": 27.080670926517573, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 33905 + }, + { + "epoch": 27.0814696485623, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 33906 + }, + { + "epoch": 27.08226837060703, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 33907 + }, + { + "epoch": 27.083067092651756, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 33908 + }, + { + "epoch": 27.083865814696484, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 33909 + }, + { + "epoch": 27.084664536741215, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 33910 + }, + { + "epoch": 27.085463258785943, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 33911 + }, + { + "epoch": 27.08626198083067, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 33912 + }, + { + "epoch": 27.087060702875398, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.072, + "step": 33913 + }, + { + "epoch": 27.08785942492013, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 33914 + }, + { + "epoch": 27.088658146964857, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 33915 + }, + { + "epoch": 27.089456869009584, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 33916 + }, + { + "epoch": 27.090255591054312, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 33917 + }, + { + "epoch": 27.09105431309904, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 33918 + }, + { + "epoch": 27.09185303514377, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.1024, + "step": 33919 + }, + { + "epoch": 27.0926517571885, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 33920 + }, + { + "epoch": 27.093450479233226, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 33921 + }, + { + "epoch": 27.094249201277954, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 33922 + }, + { + "epoch": 27.095047923322685, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 33923 + }, + { + "epoch": 27.095846645367413, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 33924 + }, + { + "epoch": 27.09664536741214, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 33925 + }, + { + "epoch": 27.097444089456868, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 33926 + }, + { + "epoch": 27.0982428115016, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 33927 + }, + { + "epoch": 27.099041533546327, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 33928 + }, + { + "epoch": 27.099840255591054, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 33929 + }, + { + "epoch": 27.100638977635782, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 33930 + }, + { + "epoch": 27.10143769968051, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 33931 + }, + { + "epoch": 27.10223642172524, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 33932 + }, + { + "epoch": 27.10303514376997, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 33933 + }, + { + "epoch": 27.103833865814696, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33934 + }, + { + "epoch": 27.104632587859424, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 33935 + }, + { + "epoch": 27.105431309904155, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 33936 + }, + { + "epoch": 27.106230031948883, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 33937 + }, + { + "epoch": 27.10702875399361, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 33938 + }, + { + "epoch": 27.107827476038338, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 33939 + }, + { + "epoch": 27.108626198083066, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 33940 + }, + { + "epoch": 27.109424920127797, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 33941 + }, + { + "epoch": 27.110223642172524, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 33942 + }, + { + "epoch": 27.111022364217252, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 33943 + }, + { + "epoch": 27.11182108626198, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 33944 + }, + { + "epoch": 27.11261980830671, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 33945 + }, + { + "epoch": 27.11341853035144, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 33946 + }, + { + "epoch": 27.114217252396166, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 33947 + }, + { + "epoch": 27.115015974440894, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 33948 + }, + { + "epoch": 27.11581469648562, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 33949 + }, + { + "epoch": 27.116613418530353, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0744, + "step": 33950 + }, + { + "epoch": 27.11741214057508, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 33951 + }, + { + "epoch": 27.118210862619808, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0773, + "step": 33952 + }, + { + "epoch": 27.119009584664536, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 33953 + }, + { + "epoch": 27.119808306709267, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 33954 + }, + { + "epoch": 27.120607028753994, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 33955 + }, + { + "epoch": 27.121405750798722, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 33956 + }, + { + "epoch": 27.12220447284345, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 33957 + }, + { + "epoch": 27.123003194888177, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 33958 + }, + { + "epoch": 27.12380191693291, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 33959 + }, + { + "epoch": 27.124600638977636, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 33960 + }, + { + "epoch": 27.125399361022364, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 33961 + }, + { + "epoch": 27.12619808306709, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 33962 + }, + { + "epoch": 27.126996805111823, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 33963 + }, + { + "epoch": 27.12779552715655, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 33964 + }, + { + "epoch": 27.128594249201278, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33965 + }, + { + "epoch": 27.129392971246006, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 33966 + }, + { + "epoch": 27.130191693290733, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 33967 + }, + { + "epoch": 27.130990415335464, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 33968 + }, + { + "epoch": 27.131789137380192, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 33969 + }, + { + "epoch": 27.13258785942492, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 33970 + }, + { + "epoch": 27.133386581469647, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 33971 + }, + { + "epoch": 27.13418530351438, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 33972 + }, + { + "epoch": 27.134984025559106, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 33973 + }, + { + "epoch": 27.135782747603834, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 33974 + }, + { + "epoch": 27.13658146964856, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 33975 + }, + { + "epoch": 27.13738019169329, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 33976 + }, + { + "epoch": 27.13817891373802, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 33977 + }, + { + "epoch": 27.138977635782748, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 33978 + }, + { + "epoch": 27.139776357827476, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 33979 + }, + { + "epoch": 27.140575079872203, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 33980 + }, + { + "epoch": 27.141373801916934, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 33981 + }, + { + "epoch": 27.142172523961662, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 33982 + }, + { + "epoch": 27.14297124600639, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 33983 + }, + { + "epoch": 27.143769968051117, + "grad_norm": 0.296875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 33984 + }, + { + "epoch": 27.144568690095845, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 33985 + }, + { + "epoch": 27.145367412140576, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 33986 + }, + { + "epoch": 27.146166134185304, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 33987 + }, + { + "epoch": 27.14696485623003, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 33988 + }, + { + "epoch": 27.14776357827476, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 33989 + }, + { + "epoch": 27.14856230031949, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 33990 + }, + { + "epoch": 27.149361022364218, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 33991 + }, + { + "epoch": 27.150159744408946, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 33992 + }, + { + "epoch": 27.150958466453673, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 33993 + }, + { + "epoch": 27.1517571884984, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 33994 + }, + { + "epoch": 27.152555910543132, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 33995 + }, + { + "epoch": 27.15335463258786, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 33996 + }, + { + "epoch": 27.154153354632587, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 33997 + }, + { + "epoch": 27.154952076677315, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 33998 + }, + { + "epoch": 27.155750798722046, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 33999 + }, + { + "epoch": 27.156549520766774, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 34000 + }, + { + "epoch": 27.1573482428115, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 34001 + }, + { + "epoch": 27.15814696485623, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 34002 + }, + { + "epoch": 27.15894568690096, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34003 + }, + { + "epoch": 27.159744408945688, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 34004 + }, + { + "epoch": 27.160543130990416, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 34005 + }, + { + "epoch": 27.161341853035143, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 34006 + }, + { + "epoch": 27.16214057507987, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0726, + "step": 34007 + }, + { + "epoch": 27.162939297124602, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 34008 + }, + { + "epoch": 27.16373801916933, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 34009 + }, + { + "epoch": 27.164536741214057, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 34010 + }, + { + "epoch": 27.165335463258785, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 34011 + }, + { + "epoch": 27.166134185303516, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 34012 + }, + { + "epoch": 27.166932907348244, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 34013 + }, + { + "epoch": 27.16773162939297, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 34014 + }, + { + "epoch": 27.1685303514377, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 34015 + }, + { + "epoch": 27.169329073482427, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 34016 + }, + { + "epoch": 27.170127795527158, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 34017 + }, + { + "epoch": 27.170926517571885, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 34018 + }, + { + "epoch": 27.171725239616613, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 34019 + }, + { + "epoch": 27.17252396166134, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 34020 + }, + { + "epoch": 27.173322683706072, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 34021 + }, + { + "epoch": 27.1741214057508, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 34022 + }, + { + "epoch": 27.174920127795527, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 34023 + }, + { + "epoch": 27.175718849840255, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 34024 + }, + { + "epoch": 27.176517571884983, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 34025 + }, + { + "epoch": 27.177316293929714, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 34026 + }, + { + "epoch": 27.17811501597444, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 34027 + }, + { + "epoch": 27.17891373801917, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34028 + }, + { + "epoch": 27.179712460063897, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 34029 + }, + { + "epoch": 27.180511182108628, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 34030 + }, + { + "epoch": 27.181309904153355, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34031 + }, + { + "epoch": 27.182108626198083, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 34032 + }, + { + "epoch": 27.18290734824281, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 34033 + }, + { + "epoch": 27.18370607028754, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 34034 + }, + { + "epoch": 27.18450479233227, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 34035 + }, + { + "epoch": 27.185303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 34036 + }, + { + "epoch": 27.186102236421725, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 34037 + }, + { + "epoch": 27.186900958466452, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 34038 + }, + { + "epoch": 27.187699680511184, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0763, + "step": 34039 + }, + { + "epoch": 27.18849840255591, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 34040 + }, + { + "epoch": 27.18929712460064, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 34041 + }, + { + "epoch": 27.190095846645367, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 34042 + }, + { + "epoch": 27.190894568690094, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 34043 + }, + { + "epoch": 27.191693290734825, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 34044 + }, + { + "epoch": 27.192492012779553, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 34045 + }, + { + "epoch": 27.19329073482428, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 34046 + }, + { + "epoch": 27.19408945686901, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 34047 + }, + { + "epoch": 27.19488817891374, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 34048 + }, + { + "epoch": 27.195686900958467, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 34049 + }, + { + "epoch": 27.196485623003195, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 34050 + }, + { + "epoch": 27.197284345047922, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 34051 + }, + { + "epoch": 27.19808306709265, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 34052 + }, + { + "epoch": 27.19888178913738, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34053 + }, + { + "epoch": 27.19968051118211, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 34054 + }, + { + "epoch": 27.200479233226837, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 34055 + }, + { + "epoch": 27.201277955271564, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 34056 + }, + { + "epoch": 27.202076677316295, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 34057 + }, + { + "epoch": 27.202875399361023, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34058 + }, + { + "epoch": 27.20367412140575, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 34059 + }, + { + "epoch": 27.20447284345048, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 34060 + }, + { + "epoch": 27.205271565495206, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 34061 + }, + { + "epoch": 27.206070287539937, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 34062 + }, + { + "epoch": 27.206869009584665, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 34063 + }, + { + "epoch": 27.207667731629392, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 34064 + }, + { + "epoch": 27.20846645367412, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 34065 + }, + { + "epoch": 27.20926517571885, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 34066 + }, + { + "epoch": 27.21006389776358, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 34067 + }, + { + "epoch": 27.210862619808307, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 34068 + }, + { + "epoch": 27.211661341853034, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 34069 + }, + { + "epoch": 27.212460063897762, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 34070 + }, + { + "epoch": 27.213258785942493, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34071 + }, + { + "epoch": 27.21405750798722, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 34072 + }, + { + "epoch": 27.21485623003195, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 34073 + }, + { + "epoch": 27.215654952076676, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 34074 + }, + { + "epoch": 27.216453674121407, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 34075 + }, + { + "epoch": 27.217252396166135, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 34076 + }, + { + "epoch": 27.218051118210862, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 34077 + }, + { + "epoch": 27.21884984025559, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 34078 + }, + { + "epoch": 27.21964856230032, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 34079 + }, + { + "epoch": 27.22044728434505, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 34080 + }, + { + "epoch": 27.221246006389777, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34081 + }, + { + "epoch": 27.222044728434504, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 34082 + }, + { + "epoch": 27.222843450479232, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34083 + }, + { + "epoch": 27.223642172523963, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 34084 + }, + { + "epoch": 27.22444089456869, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 34085 + }, + { + "epoch": 27.22523961661342, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 34086 + }, + { + "epoch": 27.226038338658146, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 34087 + }, + { + "epoch": 27.226837060702877, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 34088 + }, + { + "epoch": 27.227635782747605, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 34089 + }, + { + "epoch": 27.228434504792332, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 34090 + }, + { + "epoch": 27.22923322683706, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 34091 + }, + { + "epoch": 27.230031948881788, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 34092 + }, + { + "epoch": 27.23083067092652, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 34093 + }, + { + "epoch": 27.231629392971247, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 34094 + }, + { + "epoch": 27.232428115015974, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 34095 + }, + { + "epoch": 27.233226837060702, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 34096 + }, + { + "epoch": 27.234025559105433, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 34097 + }, + { + "epoch": 27.23482428115016, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 34098 + }, + { + "epoch": 27.23562300319489, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 34099 + }, + { + "epoch": 27.236421725239616, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 34100 + }, + { + "epoch": 27.237220447284344, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 34101 + }, + { + "epoch": 27.238019169329075, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 34102 + }, + { + "epoch": 27.238817891373802, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 34103 + }, + { + "epoch": 27.23961661341853, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 34104 + }, + { + "epoch": 27.240415335463258, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 34105 + }, + { + "epoch": 27.24121405750799, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 34106 + }, + { + "epoch": 27.242012779552716, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 34107 + }, + { + "epoch": 27.242811501597444, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 34108 + }, + { + "epoch": 27.24361022364217, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 34109 + }, + { + "epoch": 27.2444089456869, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 34110 + }, + { + "epoch": 27.24520766773163, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 34111 + }, + { + "epoch": 27.24600638977636, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 34112 + }, + { + "epoch": 27.246805111821086, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 34113 + }, + { + "epoch": 27.247603833865814, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 34114 + }, + { + "epoch": 27.248402555910545, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 34115 + }, + { + "epoch": 27.249201277955272, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 34116 + }, + { + "epoch": 27.25, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 34117 + }, + { + "epoch": 27.250798722044728, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 34118 + }, + { + "epoch": 27.251597444089455, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 34119 + }, + { + "epoch": 27.252396166134186, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 34120 + }, + { + "epoch": 27.253194888178914, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 34121 + }, + { + "epoch": 27.25399361022364, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 34122 + }, + { + "epoch": 27.25479233226837, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 34123 + }, + { + "epoch": 27.2555910543131, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 34124 + }, + { + "epoch": 27.25638977635783, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 34125 + }, + { + "epoch": 27.257188498402556, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 34126 + }, + { + "epoch": 27.257987220447284, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 34127 + }, + { + "epoch": 27.25878594249201, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 34128 + }, + { + "epoch": 27.259584664536742, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 34129 + }, + { + "epoch": 27.26038338658147, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 34130 + }, + { + "epoch": 27.261182108626198, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 34131 + }, + { + "epoch": 27.261980830670925, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 34132 + }, + { + "epoch": 27.262779552715656, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 34133 + }, + { + "epoch": 27.263578274760384, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 34134 + }, + { + "epoch": 27.26437699680511, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 34135 + }, + { + "epoch": 27.26517571884984, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 34136 + }, + { + "epoch": 27.265974440894567, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 34137 + }, + { + "epoch": 27.266773162939298, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 34138 + }, + { + "epoch": 27.267571884984026, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 34139 + }, + { + "epoch": 27.268370607028753, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 34140 + }, + { + "epoch": 27.26916932907348, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 34141 + }, + { + "epoch": 27.269968051118212, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 34142 + }, + { + "epoch": 27.27076677316294, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 34143 + }, + { + "epoch": 27.271565495207668, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34144 + }, + { + "epoch": 27.272364217252395, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 34145 + }, + { + "epoch": 27.273162939297123, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34146 + }, + { + "epoch": 27.273961661341854, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 34147 + }, + { + "epoch": 27.27476038338658, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 34148 + }, + { + "epoch": 27.27555910543131, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 34149 + }, + { + "epoch": 27.276357827476037, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 34150 + }, + { + "epoch": 27.277156549520768, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 34151 + }, + { + "epoch": 27.277955271565496, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 34152 + }, + { + "epoch": 27.278753993610223, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 34153 + }, + { + "epoch": 27.27955271565495, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 34154 + }, + { + "epoch": 27.28035143769968, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 34155 + }, + { + "epoch": 27.28115015974441, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 34156 + }, + { + "epoch": 27.281948881789138, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 34157 + }, + { + "epoch": 27.282747603833865, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 34158 + }, + { + "epoch": 27.283546325878593, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 34159 + }, + { + "epoch": 27.284345047923324, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 34160 + }, + { + "epoch": 27.28514376996805, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 34161 + }, + { + "epoch": 27.28594249201278, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 34162 + }, + { + "epoch": 27.286741214057507, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 34163 + }, + { + "epoch": 27.287539936102238, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 34164 + }, + { + "epoch": 27.288338658146966, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 34165 + }, + { + "epoch": 27.289137380191693, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 34166 + }, + { + "epoch": 27.28993610223642, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 34167 + }, + { + "epoch": 27.29073482428115, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 34168 + }, + { + "epoch": 27.29153354632588, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 34169 + }, + { + "epoch": 27.292332268370608, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34170 + }, + { + "epoch": 27.293130990415335, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 34171 + }, + { + "epoch": 27.293929712460063, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 34172 + }, + { + "epoch": 27.294728434504794, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34173 + }, + { + "epoch": 27.29552715654952, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34174 + }, + { + "epoch": 27.29632587859425, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 34175 + }, + { + "epoch": 27.297124600638977, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34176 + }, + { + "epoch": 27.297923322683705, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 34177 + }, + { + "epoch": 27.298722044728436, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 34178 + }, + { + "epoch": 27.299520766773163, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 34179 + }, + { + "epoch": 27.30031948881789, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 34180 + }, + { + "epoch": 27.30111821086262, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 34181 + }, + { + "epoch": 27.30191693290735, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 34182 + }, + { + "epoch": 27.302715654952078, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 34183 + }, + { + "epoch": 27.303514376996805, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34184 + }, + { + "epoch": 27.304313099041533, + "grad_norm": 0.2421875, + "learning_rate": 0.0001, + "loss": 1.074, + "step": 34185 + }, + { + "epoch": 27.30511182108626, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34186 + }, + { + "epoch": 27.30591054313099, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 34187 + }, + { + "epoch": 27.30670926517572, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0763, + "step": 34188 + }, + { + "epoch": 27.307507987220447, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 34189 + }, + { + "epoch": 27.308306709265175, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 34190 + }, + { + "epoch": 27.309105431309906, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 34191 + }, + { + "epoch": 27.309904153354633, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 34192 + }, + { + "epoch": 27.31070287539936, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 34193 + }, + { + "epoch": 27.31150159744409, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0759, + "step": 34194 + }, + { + "epoch": 27.312300319488816, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 34195 + }, + { + "epoch": 27.313099041533548, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 34196 + }, + { + "epoch": 27.313897763578275, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34197 + }, + { + "epoch": 27.314696485623003, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 34198 + }, + { + "epoch": 27.31549520766773, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 34199 + }, + { + "epoch": 27.31629392971246, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 34200 + }, + { + "epoch": 27.31709265175719, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 34201 + }, + { + "epoch": 27.317891373801917, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 34202 + }, + { + "epoch": 27.318690095846645, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 34203 + }, + { + "epoch": 27.319488817891372, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 34204 + }, + { + "epoch": 27.320287539936103, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 34205 + }, + { + "epoch": 27.32108626198083, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 34206 + }, + { + "epoch": 27.32188498402556, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 34207 + }, + { + "epoch": 27.322683706070286, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 34208 + }, + { + "epoch": 27.323482428115017, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34209 + }, + { + "epoch": 27.324281150159745, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 34210 + }, + { + "epoch": 27.325079872204473, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 34211 + }, + { + "epoch": 27.3258785942492, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 34212 + }, + { + "epoch": 27.326677316293928, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 34213 + }, + { + "epoch": 27.32747603833866, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 34214 + }, + { + "epoch": 27.328274760383387, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 34215 + }, + { + "epoch": 27.329073482428115, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.1045, + "step": 34216 + }, + { + "epoch": 27.329872204472842, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 34217 + }, + { + "epoch": 27.330670926517573, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 34218 + }, + { + "epoch": 27.3314696485623, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 34219 + }, + { + "epoch": 27.33226837060703, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 34220 + }, + { + "epoch": 27.333067092651756, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 34221 + }, + { + "epoch": 27.333865814696484, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 34222 + }, + { + "epoch": 27.334664536741215, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 34223 + }, + { + "epoch": 27.335463258785943, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 34224 + }, + { + "epoch": 27.33626198083067, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 34225 + }, + { + "epoch": 27.337060702875398, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 34226 + }, + { + "epoch": 27.33785942492013, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 34227 + }, + { + "epoch": 27.338658146964857, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 34228 + }, + { + "epoch": 27.339456869009584, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 34229 + }, + { + "epoch": 27.340255591054312, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 34230 + }, + { + "epoch": 27.341054313099043, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 34231 + }, + { + "epoch": 27.34185303514377, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 34232 + }, + { + "epoch": 27.3426517571885, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 34233 + }, + { + "epoch": 27.343450479233226, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34234 + }, + { + "epoch": 27.344249201277954, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 34235 + }, + { + "epoch": 27.345047923322685, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 34236 + }, + { + "epoch": 27.345846645367413, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 34237 + }, + { + "epoch": 27.34664536741214, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 34238 + }, + { + "epoch": 27.347444089456868, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 34239 + }, + { + "epoch": 27.3482428115016, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 34240 + }, + { + "epoch": 27.349041533546327, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 34241 + }, + { + "epoch": 27.349840255591054, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 34242 + }, + { + "epoch": 27.350638977635782, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 34243 + }, + { + "epoch": 27.35143769968051, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 34244 + }, + { + "epoch": 27.35223642172524, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 34245 + }, + { + "epoch": 27.35303514376997, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 34246 + }, + { + "epoch": 27.353833865814696, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 34247 + }, + { + "epoch": 27.354632587859424, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 34248 + }, + { + "epoch": 27.355431309904155, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 34249 + }, + { + "epoch": 27.356230031948883, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 34250 + }, + { + "epoch": 27.35702875399361, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 34251 + }, + { + "epoch": 27.357827476038338, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 34252 + }, + { + "epoch": 27.358626198083066, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 34253 + }, + { + "epoch": 27.359424920127797, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 34254 + }, + { + "epoch": 27.360223642172524, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 34255 + }, + { + "epoch": 27.361022364217252, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 34256 + }, + { + "epoch": 27.36182108626198, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 34257 + }, + { + "epoch": 27.36261980830671, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 34258 + }, + { + "epoch": 27.36341853035144, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 34259 + }, + { + "epoch": 27.364217252396166, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 34260 + }, + { + "epoch": 27.365015974440894, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34261 + }, + { + "epoch": 27.36581469648562, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 34262 + }, + { + "epoch": 27.366613418530353, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34263 + }, + { + "epoch": 27.36741214057508, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 34264 + }, + { + "epoch": 27.368210862619808, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 34265 + }, + { + "epoch": 27.369009584664536, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 34266 + }, + { + "epoch": 27.369808306709267, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 34267 + }, + { + "epoch": 27.370607028753994, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 34268 + }, + { + "epoch": 27.371405750798722, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 34269 + }, + { + "epoch": 27.37220447284345, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 34270 + }, + { + "epoch": 27.373003194888177, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 34271 + }, + { + "epoch": 27.37380191693291, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 34272 + }, + { + "epoch": 27.374600638977636, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 34273 + }, + { + "epoch": 27.375399361022364, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 34274 + }, + { + "epoch": 27.37619808306709, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 34275 + }, + { + "epoch": 27.376996805111823, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 34276 + }, + { + "epoch": 27.37779552715655, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 34277 + }, + { + "epoch": 27.378594249201278, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 34278 + }, + { + "epoch": 27.379392971246006, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 34279 + }, + { + "epoch": 27.380191693290733, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34280 + }, + { + "epoch": 27.380990415335464, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 34281 + }, + { + "epoch": 27.381789137380192, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 34282 + }, + { + "epoch": 27.38258785942492, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 34283 + }, + { + "epoch": 27.383386581469647, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 34284 + }, + { + "epoch": 27.38418530351438, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 34285 + }, + { + "epoch": 27.384984025559106, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 34286 + }, + { + "epoch": 27.385782747603834, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 34287 + }, + { + "epoch": 27.38658146964856, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 34288 + }, + { + "epoch": 27.38738019169329, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 34289 + }, + { + "epoch": 27.38817891373802, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 34290 + }, + { + "epoch": 27.388977635782748, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 34291 + }, + { + "epoch": 27.389776357827476, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 34292 + }, + { + "epoch": 27.390575079872203, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 34293 + }, + { + "epoch": 27.391373801916934, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 34294 + }, + { + "epoch": 27.392172523961662, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 34295 + }, + { + "epoch": 27.39297124600639, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 34296 + }, + { + "epoch": 27.393769968051117, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 34297 + }, + { + "epoch": 27.394568690095845, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 34298 + }, + { + "epoch": 27.395367412140576, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 34299 + }, + { + "epoch": 27.396166134185304, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 34300 + }, + { + "epoch": 27.39696485623003, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 34301 + }, + { + "epoch": 27.39776357827476, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 34302 + }, + { + "epoch": 27.39856230031949, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 34303 + }, + { + "epoch": 27.399361022364218, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 34304 + }, + { + "epoch": 27.400159744408946, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 34305 + }, + { + "epoch": 27.400958466453673, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 34306 + }, + { + "epoch": 27.4017571884984, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 34307 + }, + { + "epoch": 27.402555910543132, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 34308 + }, + { + "epoch": 27.40335463258786, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 34309 + }, + { + "epoch": 27.404153354632587, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34310 + }, + { + "epoch": 27.404952076677315, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 34311 + }, + { + "epoch": 27.405750798722046, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 34312 + }, + { + "epoch": 27.406549520766774, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 34313 + }, + { + "epoch": 27.4073482428115, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 34314 + }, + { + "epoch": 27.40814696485623, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 34315 + }, + { + "epoch": 27.408945686900957, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 34316 + }, + { + "epoch": 27.409744408945688, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 34317 + }, + { + "epoch": 27.410543130990416, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 34318 + }, + { + "epoch": 27.411341853035143, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 34319 + }, + { + "epoch": 27.41214057507987, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 34320 + }, + { + "epoch": 27.412939297124602, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 34321 + }, + { + "epoch": 27.41373801916933, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 34322 + }, + { + "epoch": 27.414536741214057, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 34323 + }, + { + "epoch": 27.415335463258785, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 34324 + }, + { + "epoch": 27.416134185303516, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 34325 + }, + { + "epoch": 27.416932907348244, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 34326 + }, + { + "epoch": 27.41773162939297, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 34327 + }, + { + "epoch": 27.4185303514377, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 34328 + }, + { + "epoch": 27.419329073482427, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 34329 + }, + { + "epoch": 27.420127795527158, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34330 + }, + { + "epoch": 27.420926517571885, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 34331 + }, + { + "epoch": 27.421725239616613, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 34332 + }, + { + "epoch": 27.42252396166134, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 34333 + }, + { + "epoch": 27.423322683706072, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 34334 + }, + { + "epoch": 27.4241214057508, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 34335 + }, + { + "epoch": 27.424920127795527, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 34336 + }, + { + "epoch": 27.425718849840255, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 34337 + }, + { + "epoch": 27.426517571884983, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 34338 + }, + { + "epoch": 27.427316293929714, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 34339 + }, + { + "epoch": 27.42811501597444, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 34340 + }, + { + "epoch": 27.42891373801917, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 34341 + }, + { + "epoch": 27.429712460063897, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 34342 + }, + { + "epoch": 27.430511182108628, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 34343 + }, + { + "epoch": 27.431309904153355, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 34344 + }, + { + "epoch": 27.432108626198083, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 34345 + }, + { + "epoch": 27.43290734824281, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 34346 + }, + { + "epoch": 27.43370607028754, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 34347 + }, + { + "epoch": 27.43450479233227, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 34348 + }, + { + "epoch": 27.435303514376997, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 34349 + }, + { + "epoch": 27.436102236421725, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 34350 + }, + { + "epoch": 27.436900958466452, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 34351 + }, + { + "epoch": 27.437699680511184, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 34352 + }, + { + "epoch": 27.43849840255591, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 34353 + }, + { + "epoch": 27.43929712460064, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 34354 + }, + { + "epoch": 27.440095846645367, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 34355 + }, + { + "epoch": 27.440894568690094, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 34356 + }, + { + "epoch": 27.441693290734825, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 34357 + }, + { + "epoch": 27.442492012779553, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 34358 + }, + { + "epoch": 27.44329073482428, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 34359 + }, + { + "epoch": 27.44408945686901, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 34360 + }, + { + "epoch": 27.44488817891374, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34361 + }, + { + "epoch": 27.445686900958467, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 34362 + }, + { + "epoch": 27.446485623003195, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 34363 + }, + { + "epoch": 27.447284345047922, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 34364 + }, + { + "epoch": 27.44808306709265, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 34365 + }, + { + "epoch": 27.44888178913738, + "grad_norm": 0.09521484375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 34366 + }, + { + "epoch": 27.44968051118211, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 34367 + }, + { + "epoch": 27.450479233226837, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 34368 + }, + { + "epoch": 27.451277955271564, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 34369 + }, + { + "epoch": 27.452076677316295, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 34370 + }, + { + "epoch": 27.452875399361023, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 34371 + }, + { + "epoch": 27.45367412140575, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34372 + }, + { + "epoch": 27.45447284345048, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34373 + }, + { + "epoch": 27.455271565495206, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 34374 + }, + { + "epoch": 27.456070287539937, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 34375 + }, + { + "epoch": 27.456869009584665, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 34376 + }, + { + "epoch": 27.457667731629392, + "grad_norm": 0.09619140625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 34377 + }, + { + "epoch": 27.45846645367412, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 34378 + }, + { + "epoch": 27.45926517571885, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 34379 + }, + { + "epoch": 27.46006389776358, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 34380 + }, + { + "epoch": 27.460862619808307, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 34381 + }, + { + "epoch": 27.461661341853034, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 34382 + }, + { + "epoch": 27.462460063897762, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 34383 + }, + { + "epoch": 27.463258785942493, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 34384 + }, + { + "epoch": 27.46405750798722, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 34385 + }, + { + "epoch": 27.46485623003195, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 34386 + }, + { + "epoch": 27.465654952076676, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 34387 + }, + { + "epoch": 27.466453674121407, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 34388 + }, + { + "epoch": 27.467252396166135, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 34389 + }, + { + "epoch": 27.468051118210862, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 34390 + }, + { + "epoch": 27.46884984025559, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 34391 + }, + { + "epoch": 27.46964856230032, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 34392 + }, + { + "epoch": 27.47044728434505, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 34393 + }, + { + "epoch": 27.471246006389777, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 34394 + }, + { + "epoch": 27.472044728434504, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 34395 + }, + { + "epoch": 27.472843450479232, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 34396 + }, + { + "epoch": 27.473642172523963, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34397 + }, + { + "epoch": 27.47444089456869, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 34398 + }, + { + "epoch": 27.47523961661342, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 34399 + }, + { + "epoch": 27.476038338658146, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 34400 + }, + { + "epoch": 27.476837060702877, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 34401 + }, + { + "epoch": 27.477635782747605, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 34402 + }, + { + "epoch": 27.478434504792332, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 34403 + }, + { + "epoch": 27.47923322683706, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0762, + "step": 34404 + }, + { + "epoch": 27.480031948881788, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 34405 + }, + { + "epoch": 27.48083067092652, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 34406 + }, + { + "epoch": 27.481629392971247, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 34407 + }, + { + "epoch": 27.482428115015974, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 34408 + }, + { + "epoch": 27.483226837060702, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 34409 + }, + { + "epoch": 27.484025559105433, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 34410 + }, + { + "epoch": 27.48482428115016, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 34411 + }, + { + "epoch": 27.48562300319489, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 34412 + }, + { + "epoch": 27.486421725239616, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 34413 + }, + { + "epoch": 27.487220447284344, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 34414 + }, + { + "epoch": 27.488019169329075, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 34415 + }, + { + "epoch": 27.488817891373802, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 34416 + }, + { + "epoch": 27.48961661341853, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 34417 + }, + { + "epoch": 27.490415335463258, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 34418 + }, + { + "epoch": 27.49121405750799, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 34419 + }, + { + "epoch": 27.492012779552716, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34420 + }, + { + "epoch": 27.492811501597444, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 34421 + }, + { + "epoch": 27.49361022364217, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 34422 + }, + { + "epoch": 27.4944089456869, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 34423 + }, + { + "epoch": 27.49520766773163, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 34424 + }, + { + "epoch": 27.49600638977636, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 34425 + }, + { + "epoch": 27.496805111821086, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34426 + }, + { + "epoch": 27.497603833865814, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 34427 + }, + { + "epoch": 27.498402555910545, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 34428 + }, + { + "epoch": 27.499201277955272, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 34429 + }, + { + "epoch": 27.5, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 34430 + }, + { + "epoch": 27.500798722044728, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 34431 + }, + { + "epoch": 27.501597444089455, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34432 + }, + { + "epoch": 27.502396166134186, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 34433 + }, + { + "epoch": 27.503194888178914, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34434 + }, + { + "epoch": 27.50399361022364, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 34435 + }, + { + "epoch": 27.50479233226837, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 34436 + }, + { + "epoch": 27.5055910543131, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 34437 + }, + { + "epoch": 27.50638977635783, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 34438 + }, + { + "epoch": 27.507188498402556, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 34439 + }, + { + "epoch": 27.507987220447284, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 34440 + }, + { + "epoch": 27.50878594249201, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 34441 + }, + { + "epoch": 27.509584664536742, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 34442 + }, + { + "epoch": 27.51038338658147, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 34443 + }, + { + "epoch": 27.511182108626198, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 34444 + }, + { + "epoch": 27.511980830670925, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 34445 + }, + { + "epoch": 27.512779552715656, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34446 + }, + { + "epoch": 27.513578274760384, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 34447 + }, + { + "epoch": 27.51437699680511, + "grad_norm": 0.10009765625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 34448 + }, + { + "epoch": 27.51517571884984, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 34449 + }, + { + "epoch": 27.515974440894567, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 34450 + }, + { + "epoch": 27.516773162939298, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 34451 + }, + { + "epoch": 27.517571884984026, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 34452 + }, + { + "epoch": 27.518370607028753, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 34453 + }, + { + "epoch": 27.51916932907348, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 34454 + }, + { + "epoch": 27.519968051118212, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 34455 + }, + { + "epoch": 27.52076677316294, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 34456 + }, + { + "epoch": 27.521565495207668, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 34457 + }, + { + "epoch": 27.522364217252395, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 34458 + }, + { + "epoch": 27.523162939297123, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 34459 + }, + { + "epoch": 27.523961661341854, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 34460 + }, + { + "epoch": 27.52476038338658, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34461 + }, + { + "epoch": 27.52555910543131, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 34462 + }, + { + "epoch": 27.526357827476037, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 34463 + }, + { + "epoch": 27.527156549520768, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 34464 + }, + { + "epoch": 27.527955271565496, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 34465 + }, + { + "epoch": 27.528753993610223, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 34466 + }, + { + "epoch": 27.52955271565495, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 34467 + }, + { + "epoch": 27.53035143769968, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 34468 + }, + { + "epoch": 27.53115015974441, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 34469 + }, + { + "epoch": 27.531948881789138, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 34470 + }, + { + "epoch": 27.532747603833865, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 34471 + }, + { + "epoch": 27.533546325878593, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34472 + }, + { + "epoch": 27.534345047923324, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 34473 + }, + { + "epoch": 27.53514376996805, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 34474 + }, + { + "epoch": 27.53594249201278, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 34475 + }, + { + "epoch": 27.536741214057507, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 34476 + }, + { + "epoch": 27.537539936102235, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 34477 + }, + { + "epoch": 27.538338658146966, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 34478 + }, + { + "epoch": 27.539137380191693, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 34479 + }, + { + "epoch": 27.53993610223642, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 34480 + }, + { + "epoch": 27.54073482428115, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 34481 + }, + { + "epoch": 27.54153354632588, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34482 + }, + { + "epoch": 27.542332268370608, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 34483 + }, + { + "epoch": 27.543130990415335, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 34484 + }, + { + "epoch": 27.543929712460063, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 34485 + }, + { + "epoch": 27.544728434504794, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 34486 + }, + { + "epoch": 27.54552715654952, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0725, + "step": 34487 + }, + { + "epoch": 27.54632587859425, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34488 + }, + { + "epoch": 27.547124600638977, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 34489 + }, + { + "epoch": 27.547923322683705, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 34490 + }, + { + "epoch": 27.548722044728436, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 34491 + }, + { + "epoch": 27.549520766773163, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0759, + "step": 34492 + }, + { + "epoch": 27.55031948881789, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 34493 + }, + { + "epoch": 27.55111821086262, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 34494 + }, + { + "epoch": 27.55191693290735, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 34495 + }, + { + "epoch": 27.552715654952078, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34496 + }, + { + "epoch": 27.553514376996805, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 34497 + }, + { + "epoch": 27.554313099041533, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 34498 + }, + { + "epoch": 27.55511182108626, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34499 + }, + { + "epoch": 27.55591054313099, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 34500 + }, + { + "epoch": 27.55670926517572, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 34501 + }, + { + "epoch": 27.557507987220447, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 34502 + }, + { + "epoch": 27.558306709265175, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 34503 + }, + { + "epoch": 27.559105431309906, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 34504 + }, + { + "epoch": 27.559904153354633, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34505 + }, + { + "epoch": 27.56070287539936, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 34506 + }, + { + "epoch": 27.56150159744409, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 34507 + }, + { + "epoch": 27.562300319488816, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 34508 + }, + { + "epoch": 27.563099041533548, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34509 + }, + { + "epoch": 27.563897763578275, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 34510 + }, + { + "epoch": 27.564696485623003, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 34511 + }, + { + "epoch": 27.56549520766773, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 34512 + }, + { + "epoch": 27.56629392971246, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 34513 + }, + { + "epoch": 27.56709265175719, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34514 + }, + { + "epoch": 27.567891373801917, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 34515 + }, + { + "epoch": 27.568690095846645, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34516 + }, + { + "epoch": 27.569488817891372, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 34517 + }, + { + "epoch": 27.570287539936103, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 34518 + }, + { + "epoch": 27.57108626198083, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 34519 + }, + { + "epoch": 27.57188498402556, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 34520 + }, + { + "epoch": 27.572683706070286, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 34521 + }, + { + "epoch": 27.573482428115017, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 34522 + }, + { + "epoch": 27.574281150159745, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 34523 + }, + { + "epoch": 27.575079872204473, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 34524 + }, + { + "epoch": 27.5758785942492, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34525 + }, + { + "epoch": 27.576677316293928, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 34526 + }, + { + "epoch": 27.57747603833866, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 34527 + }, + { + "epoch": 27.578274760383387, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 34528 + }, + { + "epoch": 27.579073482428115, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 34529 + }, + { + "epoch": 27.579872204472842, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 34530 + }, + { + "epoch": 27.580670926517573, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 34531 + }, + { + "epoch": 27.5814696485623, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 34532 + }, + { + "epoch": 27.58226837060703, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 34533 + }, + { + "epoch": 27.583067092651756, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 34534 + }, + { + "epoch": 27.583865814696484, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 34535 + }, + { + "epoch": 27.584664536741215, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 34536 + }, + { + "epoch": 27.585463258785943, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 34537 + }, + { + "epoch": 27.58626198083067, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 34538 + }, + { + "epoch": 27.587060702875398, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34539 + }, + { + "epoch": 27.58785942492013, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 34540 + }, + { + "epoch": 27.588658146964857, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 34541 + }, + { + "epoch": 27.589456869009584, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 34542 + }, + { + "epoch": 27.590255591054312, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 34543 + }, + { + "epoch": 27.591054313099043, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 34544 + }, + { + "epoch": 27.59185303514377, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 34545 + }, + { + "epoch": 27.5926517571885, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 34546 + }, + { + "epoch": 27.593450479233226, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 34547 + }, + { + "epoch": 27.594249201277954, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 34548 + }, + { + "epoch": 27.595047923322685, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 34549 + }, + { + "epoch": 27.595846645367413, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 34550 + }, + { + "epoch": 27.59664536741214, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 34551 + }, + { + "epoch": 27.597444089456868, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 34552 + }, + { + "epoch": 27.5982428115016, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34553 + }, + { + "epoch": 27.599041533546327, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 34554 + }, + { + "epoch": 27.599840255591054, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 34555 + }, + { + "epoch": 27.600638977635782, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 34556 + }, + { + "epoch": 27.60143769968051, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 34557 + }, + { + "epoch": 27.60223642172524, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 34558 + }, + { + "epoch": 27.60303514376997, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 34559 + }, + { + "epoch": 27.603833865814696, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 34560 + }, + { + "epoch": 27.604632587859424, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 34561 + }, + { + "epoch": 27.605431309904155, + "grad_norm": 0.302734375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 34562 + }, + { + "epoch": 27.606230031948883, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 34563 + }, + { + "epoch": 27.60702875399361, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 34564 + }, + { + "epoch": 27.607827476038338, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 34565 + }, + { + "epoch": 27.608626198083066, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 34566 + }, + { + "epoch": 27.609424920127797, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 34567 + }, + { + "epoch": 27.610223642172524, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 34568 + }, + { + "epoch": 27.611022364217252, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 34569 + }, + { + "epoch": 27.61182108626198, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 34570 + }, + { + "epoch": 27.61261980830671, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 34571 + }, + { + "epoch": 27.61341853035144, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 34572 + }, + { + "epoch": 27.614217252396166, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 34573 + }, + { + "epoch": 27.615015974440894, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 34574 + }, + { + "epoch": 27.61581469648562, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 34575 + }, + { + "epoch": 27.616613418530353, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 34576 + }, + { + "epoch": 27.61741214057508, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 34577 + }, + { + "epoch": 27.618210862619808, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 34578 + }, + { + "epoch": 27.619009584664536, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 34579 + }, + { + "epoch": 27.619808306709267, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 34580 + }, + { + "epoch": 27.620607028753994, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 34581 + }, + { + "epoch": 27.621405750798722, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 34582 + }, + { + "epoch": 27.62220447284345, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34583 + }, + { + "epoch": 27.623003194888177, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 34584 + }, + { + "epoch": 27.62380191693291, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 34585 + }, + { + "epoch": 27.624600638977636, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 34586 + }, + { + "epoch": 27.625399361022364, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 34587 + }, + { + "epoch": 27.62619808306709, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34588 + }, + { + "epoch": 27.626996805111823, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 34589 + }, + { + "epoch": 27.62779552715655, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34590 + }, + { + "epoch": 27.628594249201278, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 34591 + }, + { + "epoch": 27.629392971246006, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 34592 + }, + { + "epoch": 27.630191693290733, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 34593 + }, + { + "epoch": 27.630990415335464, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0749, + "step": 34594 + }, + { + "epoch": 27.631789137380192, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 34595 + }, + { + "epoch": 27.63258785942492, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 34596 + }, + { + "epoch": 27.633386581469647, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 34597 + }, + { + "epoch": 27.63418530351438, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 34598 + }, + { + "epoch": 27.634984025559106, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 34599 + }, + { + "epoch": 27.635782747603834, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 34600 + }, + { + "epoch": 27.63658146964856, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 34601 + }, + { + "epoch": 27.63738019169329, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 34602 + }, + { + "epoch": 27.63817891373802, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 34603 + }, + { + "epoch": 27.638977635782748, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 34604 + }, + { + "epoch": 27.639776357827476, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 34605 + }, + { + "epoch": 27.640575079872203, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 34606 + }, + { + "epoch": 27.641373801916934, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 34607 + }, + { + "epoch": 27.642172523961662, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 34608 + }, + { + "epoch": 27.64297124600639, + "grad_norm": 0.306640625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 34609 + }, + { + "epoch": 27.643769968051117, + "grad_norm": 0.28515625, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 34610 + }, + { + "epoch": 27.644568690095845, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 34611 + }, + { + "epoch": 27.645367412140576, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 34612 + }, + { + "epoch": 27.646166134185304, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 34613 + }, + { + "epoch": 27.64696485623003, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 34614 + }, + { + "epoch": 27.64776357827476, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 34615 + }, + { + "epoch": 27.64856230031949, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 34616 + }, + { + "epoch": 27.649361022364218, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 34617 + }, + { + "epoch": 27.650159744408946, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 34618 + }, + { + "epoch": 27.650958466453673, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 34619 + }, + { + "epoch": 27.6517571884984, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 34620 + }, + { + "epoch": 27.652555910543132, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 34621 + }, + { + "epoch": 27.65335463258786, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34622 + }, + { + "epoch": 27.654153354632587, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 34623 + }, + { + "epoch": 27.654952076677315, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 34624 + }, + { + "epoch": 27.655750798722046, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 34625 + }, + { + "epoch": 27.656549520766774, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34626 + }, + { + "epoch": 27.6573482428115, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 34627 + }, + { + "epoch": 27.65814696485623, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 34628 + }, + { + "epoch": 27.658945686900957, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 34629 + }, + { + "epoch": 27.659744408945688, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 34630 + }, + { + "epoch": 27.660543130990416, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34631 + }, + { + "epoch": 27.661341853035143, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 34632 + }, + { + "epoch": 27.66214057507987, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 34633 + }, + { + "epoch": 27.662939297124602, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 34634 + }, + { + "epoch": 27.66373801916933, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 34635 + }, + { + "epoch": 27.664536741214057, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 34636 + }, + { + "epoch": 27.665335463258785, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 34637 + }, + { + "epoch": 27.666134185303516, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 34638 + }, + { + "epoch": 27.666932907348244, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 34639 + }, + { + "epoch": 27.66773162939297, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 34640 + }, + { + "epoch": 27.6685303514377, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34641 + }, + { + "epoch": 27.669329073482427, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34642 + }, + { + "epoch": 27.670127795527158, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 34643 + }, + { + "epoch": 27.670926517571885, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 34644 + }, + { + "epoch": 27.671725239616613, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 34645 + }, + { + "epoch": 27.67252396166134, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 34646 + }, + { + "epoch": 27.673322683706072, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 34647 + }, + { + "epoch": 27.6741214057508, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 34648 + }, + { + "epoch": 27.674920127795527, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 34649 + }, + { + "epoch": 27.675718849840255, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 34650 + }, + { + "epoch": 27.676517571884983, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 34651 + }, + { + "epoch": 27.677316293929714, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 34652 + }, + { + "epoch": 27.67811501597444, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 34653 + }, + { + "epoch": 27.67891373801917, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 34654 + }, + { + "epoch": 27.679712460063897, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 34655 + }, + { + "epoch": 27.680511182108628, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 34656 + }, + { + "epoch": 27.681309904153355, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 34657 + }, + { + "epoch": 27.682108626198083, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 34658 + }, + { + "epoch": 27.68290734824281, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 34659 + }, + { + "epoch": 27.68370607028754, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 34660 + }, + { + "epoch": 27.68450479233227, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 34661 + }, + { + "epoch": 27.685303514376997, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 34662 + }, + { + "epoch": 27.686102236421725, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 34663 + }, + { + "epoch": 27.686900958466452, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 34664 + }, + { + "epoch": 27.687699680511184, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 34665 + }, + { + "epoch": 27.68849840255591, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0757, + "step": 34666 + }, + { + "epoch": 27.68929712460064, + "grad_norm": 0.34765625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 34667 + }, + { + "epoch": 27.690095846645367, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 34668 + }, + { + "epoch": 27.690894568690094, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 34669 + }, + { + "epoch": 27.691693290734825, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 34670 + }, + { + "epoch": 27.692492012779553, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 34671 + }, + { + "epoch": 27.69329073482428, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 34672 + }, + { + "epoch": 27.69408945686901, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34673 + }, + { + "epoch": 27.69488817891374, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 34674 + }, + { + "epoch": 27.695686900958467, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 34675 + }, + { + "epoch": 27.696485623003195, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 34676 + }, + { + "epoch": 27.697284345047922, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 34677 + }, + { + "epoch": 27.69808306709265, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 34678 + }, + { + "epoch": 27.69888178913738, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 34679 + }, + { + "epoch": 27.69968051118211, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 34680 + }, + { + "epoch": 27.700479233226837, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 34681 + }, + { + "epoch": 27.701277955271564, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34682 + }, + { + "epoch": 27.702076677316295, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 34683 + }, + { + "epoch": 27.702875399361023, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 34684 + }, + { + "epoch": 27.70367412140575, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34685 + }, + { + "epoch": 27.70447284345048, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 34686 + }, + { + "epoch": 27.705271565495206, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 34687 + }, + { + "epoch": 27.706070287539937, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 34688 + }, + { + "epoch": 27.706869009584665, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 34689 + }, + { + "epoch": 27.707667731629392, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 34690 + }, + { + "epoch": 27.70846645367412, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 34691 + }, + { + "epoch": 27.70926517571885, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 34692 + }, + { + "epoch": 27.71006389776358, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 34693 + }, + { + "epoch": 27.710862619808307, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34694 + }, + { + "epoch": 27.711661341853034, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 34695 + }, + { + "epoch": 27.712460063897765, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 34696 + }, + { + "epoch": 27.713258785942493, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 34697 + }, + { + "epoch": 27.71405750798722, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 34698 + }, + { + "epoch": 27.71485623003195, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 34699 + }, + { + "epoch": 27.715654952076676, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 34700 + }, + { + "epoch": 27.716453674121407, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 34701 + }, + { + "epoch": 27.717252396166135, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 34702 + }, + { + "epoch": 27.718051118210862, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 34703 + }, + { + "epoch": 27.71884984025559, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 34704 + }, + { + "epoch": 27.71964856230032, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 34705 + }, + { + "epoch": 27.72044728434505, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 34706 + }, + { + "epoch": 27.721246006389777, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 34707 + }, + { + "epoch": 27.722044728434504, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 34708 + }, + { + "epoch": 27.722843450479232, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 34709 + }, + { + "epoch": 27.723642172523963, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 34710 + }, + { + "epoch": 27.72444089456869, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 34711 + }, + { + "epoch": 27.72523961661342, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 34712 + }, + { + "epoch": 27.726038338658146, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 34713 + }, + { + "epoch": 27.726837060702877, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34714 + }, + { + "epoch": 27.727635782747605, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 34715 + }, + { + "epoch": 27.728434504792332, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 34716 + }, + { + "epoch": 27.72923322683706, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 34717 + }, + { + "epoch": 27.730031948881788, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 34718 + }, + { + "epoch": 27.73083067092652, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 34719 + }, + { + "epoch": 27.731629392971247, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 34720 + }, + { + "epoch": 27.732428115015974, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34721 + }, + { + "epoch": 27.733226837060702, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 34722 + }, + { + "epoch": 27.734025559105433, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 34723 + }, + { + "epoch": 27.73482428115016, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 34724 + }, + { + "epoch": 27.73562300319489, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 34725 + }, + { + "epoch": 27.736421725239616, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 34726 + }, + { + "epoch": 27.737220447284344, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 34727 + }, + { + "epoch": 27.738019169329075, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 34728 + }, + { + "epoch": 27.738817891373802, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0735, + "step": 34729 + }, + { + "epoch": 27.73961661341853, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 34730 + }, + { + "epoch": 27.740415335463258, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 34731 + }, + { + "epoch": 27.74121405750799, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 34732 + }, + { + "epoch": 27.742012779552716, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 34733 + }, + { + "epoch": 27.742811501597444, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 34734 + }, + { + "epoch": 27.74361022364217, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 34735 + }, + { + "epoch": 27.7444089456869, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 34736 + }, + { + "epoch": 27.74520766773163, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 34737 + }, + { + "epoch": 27.74600638977636, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 34738 + }, + { + "epoch": 27.746805111821086, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 34739 + }, + { + "epoch": 27.747603833865814, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 34740 + }, + { + "epoch": 27.748402555910545, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 34741 + }, + { + "epoch": 27.749201277955272, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 34742 + }, + { + "epoch": 27.75, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 34743 + }, + { + "epoch": 27.750798722044728, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 34744 + }, + { + "epoch": 27.751597444089455, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 34745 + }, + { + "epoch": 27.752396166134186, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 34746 + }, + { + "epoch": 27.753194888178914, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 34747 + }, + { + "epoch": 27.75399361022364, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 34748 + }, + { + "epoch": 27.75479233226837, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 34749 + }, + { + "epoch": 27.7555910543131, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 34750 + }, + { + "epoch": 27.75638977635783, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 34751 + }, + { + "epoch": 27.757188498402556, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 34752 + }, + { + "epoch": 27.757987220447284, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 34753 + }, + { + "epoch": 27.75878594249201, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 34754 + }, + { + "epoch": 27.759584664536742, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 34755 + }, + { + "epoch": 27.76038338658147, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 34756 + }, + { + "epoch": 27.761182108626198, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 34757 + }, + { + "epoch": 27.761980830670925, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 34758 + }, + { + "epoch": 27.762779552715656, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 34759 + }, + { + "epoch": 27.763578274760384, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 34760 + }, + { + "epoch": 27.76437699680511, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 34761 + }, + { + "epoch": 27.76517571884984, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 34762 + }, + { + "epoch": 27.765974440894567, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 34763 + }, + { + "epoch": 27.766773162939298, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 34764 + }, + { + "epoch": 27.767571884984026, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 34765 + }, + { + "epoch": 27.768370607028753, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 34766 + }, + { + "epoch": 27.76916932907348, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34767 + }, + { + "epoch": 27.769968051118212, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 34768 + }, + { + "epoch": 27.77076677316294, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34769 + }, + { + "epoch": 27.771565495207668, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 34770 + }, + { + "epoch": 27.772364217252395, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 34771 + }, + { + "epoch": 27.773162939297123, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 34772 + }, + { + "epoch": 27.773961661341854, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 34773 + }, + { + "epoch": 27.77476038338658, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 34774 + }, + { + "epoch": 27.77555910543131, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 34775 + }, + { + "epoch": 27.776357827476037, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 34776 + }, + { + "epoch": 27.777156549520768, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 34777 + }, + { + "epoch": 27.777955271565496, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 34778 + }, + { + "epoch": 27.778753993610223, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 34779 + }, + { + "epoch": 27.77955271565495, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 34780 + }, + { + "epoch": 27.78035143769968, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 34781 + }, + { + "epoch": 27.78115015974441, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 34782 + }, + { + "epoch": 27.781948881789138, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34783 + }, + { + "epoch": 27.782747603833865, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 34784 + }, + { + "epoch": 27.783546325878593, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 34785 + }, + { + "epoch": 27.784345047923324, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 34786 + }, + { + "epoch": 27.78514376996805, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0733, + "step": 34787 + }, + { + "epoch": 27.78594249201278, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 34788 + }, + { + "epoch": 27.786741214057507, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 34789 + }, + { + "epoch": 27.787539936102235, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 34790 + }, + { + "epoch": 27.788338658146966, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 34791 + }, + { + "epoch": 27.789137380191693, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 34792 + }, + { + "epoch": 27.78993610223642, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 34793 + }, + { + "epoch": 27.79073482428115, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 34794 + }, + { + "epoch": 27.79153354632588, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 34795 + }, + { + "epoch": 27.792332268370608, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 34796 + }, + { + "epoch": 27.793130990415335, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 34797 + }, + { + "epoch": 27.793929712460063, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 34798 + }, + { + "epoch": 27.794728434504794, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 34799 + }, + { + "epoch": 27.79552715654952, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 34800 + }, + { + "epoch": 27.79632587859425, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 34801 + }, + { + "epoch": 27.797124600638977, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 34802 + }, + { + "epoch": 27.797923322683705, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 34803 + }, + { + "epoch": 27.798722044728436, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.101, + "step": 34804 + }, + { + "epoch": 27.799520766773163, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 34805 + }, + { + "epoch": 27.80031948881789, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 34806 + }, + { + "epoch": 27.80111821086262, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 34807 + }, + { + "epoch": 27.80191693290735, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 34808 + }, + { + "epoch": 27.802715654952078, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 34809 + }, + { + "epoch": 27.803514376996805, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 34810 + }, + { + "epoch": 27.804313099041533, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 34811 + }, + { + "epoch": 27.80511182108626, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 34812 + }, + { + "epoch": 27.80591054313099, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 34813 + }, + { + "epoch": 27.80670926517572, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 34814 + }, + { + "epoch": 27.807507987220447, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 34815 + }, + { + "epoch": 27.808306709265175, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 34816 + }, + { + "epoch": 27.809105431309906, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 34817 + }, + { + "epoch": 27.809904153354633, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 34818 + }, + { + "epoch": 27.81070287539936, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 34819 + }, + { + "epoch": 27.81150159744409, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34820 + }, + { + "epoch": 27.812300319488816, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 34821 + }, + { + "epoch": 27.813099041533548, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 34822 + }, + { + "epoch": 27.813897763578275, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 34823 + }, + { + "epoch": 27.814696485623003, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 34824 + }, + { + "epoch": 27.81549520766773, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 34825 + }, + { + "epoch": 27.81629392971246, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 34826 + }, + { + "epoch": 27.81709265175719, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 34827 + }, + { + "epoch": 27.817891373801917, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 34828 + }, + { + "epoch": 27.818690095846645, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 34829 + }, + { + "epoch": 27.819488817891372, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 34830 + }, + { + "epoch": 27.820287539936103, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 34831 + }, + { + "epoch": 27.82108626198083, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0765, + "step": 34832 + }, + { + "epoch": 27.82188498402556, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 34833 + }, + { + "epoch": 27.822683706070286, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 34834 + }, + { + "epoch": 27.823482428115017, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 34835 + }, + { + "epoch": 27.824281150159745, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 34836 + }, + { + "epoch": 27.825079872204473, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 34837 + }, + { + "epoch": 27.8258785942492, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 34838 + }, + { + "epoch": 27.826677316293928, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 34839 + }, + { + "epoch": 27.82747603833866, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 34840 + }, + { + "epoch": 27.828274760383387, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 34841 + }, + { + "epoch": 27.829073482428115, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 34842 + }, + { + "epoch": 27.829872204472842, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34843 + }, + { + "epoch": 27.830670926517573, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 34844 + }, + { + "epoch": 27.8314696485623, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 34845 + }, + { + "epoch": 27.83226837060703, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 34846 + }, + { + "epoch": 27.833067092651756, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 34847 + }, + { + "epoch": 27.833865814696484, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 34848 + }, + { + "epoch": 27.834664536741215, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 34849 + }, + { + "epoch": 27.835463258785943, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 34850 + }, + { + "epoch": 27.83626198083067, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 34851 + }, + { + "epoch": 27.837060702875398, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 34852 + }, + { + "epoch": 27.83785942492013, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 34853 + }, + { + "epoch": 27.838658146964857, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 34854 + }, + { + "epoch": 27.839456869009584, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 34855 + }, + { + "epoch": 27.840255591054312, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 34856 + }, + { + "epoch": 27.841054313099043, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 34857 + }, + { + "epoch": 27.84185303514377, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 34858 + }, + { + "epoch": 27.8426517571885, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 34859 + }, + { + "epoch": 27.843450479233226, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 34860 + }, + { + "epoch": 27.844249201277954, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 34861 + }, + { + "epoch": 27.845047923322685, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 34862 + }, + { + "epoch": 27.845846645367413, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 34863 + }, + { + "epoch": 27.84664536741214, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 34864 + }, + { + "epoch": 27.847444089456868, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 34865 + }, + { + "epoch": 27.8482428115016, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 34866 + }, + { + "epoch": 27.849041533546327, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 34867 + }, + { + "epoch": 27.849840255591054, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 34868 + }, + { + "epoch": 27.850638977635782, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 34869 + }, + { + "epoch": 27.85143769968051, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 34870 + }, + { + "epoch": 27.85223642172524, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 34871 + }, + { + "epoch": 27.85303514376997, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 34872 + }, + { + "epoch": 27.853833865814696, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 34873 + }, + { + "epoch": 27.854632587859424, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 34874 + }, + { + "epoch": 27.855431309904155, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0747, + "step": 34875 + }, + { + "epoch": 27.856230031948883, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 34876 + }, + { + "epoch": 27.85702875399361, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 34877 + }, + { + "epoch": 27.857827476038338, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 34878 + }, + { + "epoch": 27.858626198083066, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 34879 + }, + { + "epoch": 27.859424920127797, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 34880 + }, + { + "epoch": 27.860223642172524, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 34881 + }, + { + "epoch": 27.861022364217252, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 34882 + }, + { + "epoch": 27.86182108626198, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 34883 + }, + { + "epoch": 27.86261980830671, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 34884 + }, + { + "epoch": 27.86341853035144, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 34885 + }, + { + "epoch": 27.864217252396166, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 34886 + }, + { + "epoch": 27.865015974440894, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 34887 + }, + { + "epoch": 27.86581469648562, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 34888 + }, + { + "epoch": 27.866613418530353, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34889 + }, + { + "epoch": 27.86741214057508, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 34890 + }, + { + "epoch": 27.868210862619808, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 34891 + }, + { + "epoch": 27.869009584664536, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 34892 + }, + { + "epoch": 27.869808306709267, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 34893 + }, + { + "epoch": 27.870607028753994, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 34894 + }, + { + "epoch": 27.871405750798722, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 34895 + }, + { + "epoch": 27.87220447284345, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 34896 + }, + { + "epoch": 27.873003194888177, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 34897 + }, + { + "epoch": 27.87380191693291, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 34898 + }, + { + "epoch": 27.874600638977636, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 34899 + }, + { + "epoch": 27.875399361022364, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 34900 + }, + { + "epoch": 27.87619808306709, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34901 + }, + { + "epoch": 27.876996805111823, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 34902 + }, + { + "epoch": 27.87779552715655, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 34903 + }, + { + "epoch": 27.878594249201278, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 34904 + }, + { + "epoch": 27.879392971246006, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 34905 + }, + { + "epoch": 27.880191693290733, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 34906 + }, + { + "epoch": 27.880990415335464, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 34907 + }, + { + "epoch": 27.881789137380192, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 34908 + }, + { + "epoch": 27.88258785942492, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 34909 + }, + { + "epoch": 27.883386581469647, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 34910 + }, + { + "epoch": 27.88418530351438, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 34911 + }, + { + "epoch": 27.884984025559106, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 34912 + }, + { + "epoch": 27.885782747603834, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 34913 + }, + { + "epoch": 27.88658146964856, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 34914 + }, + { + "epoch": 27.88738019169329, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 34915 + }, + { + "epoch": 27.88817891373802, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 34916 + }, + { + "epoch": 27.888977635782748, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 34917 + }, + { + "epoch": 27.889776357827476, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 34918 + }, + { + "epoch": 27.890575079872203, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 34919 + }, + { + "epoch": 27.891373801916934, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 34920 + }, + { + "epoch": 27.892172523961662, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 34921 + }, + { + "epoch": 27.89297124600639, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 34922 + }, + { + "epoch": 27.893769968051117, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 34923 + }, + { + "epoch": 27.894568690095845, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 34924 + }, + { + "epoch": 27.895367412140576, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 34925 + }, + { + "epoch": 27.896166134185304, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 34926 + }, + { + "epoch": 27.89696485623003, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 34927 + }, + { + "epoch": 27.89776357827476, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 34928 + }, + { + "epoch": 27.89856230031949, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 34929 + }, + { + "epoch": 27.899361022364218, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 34930 + }, + { + "epoch": 27.900159744408946, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34931 + }, + { + "epoch": 27.900958466453673, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 34932 + }, + { + "epoch": 27.9017571884984, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 34933 + }, + { + "epoch": 27.902555910543132, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 34934 + }, + { + "epoch": 27.90335463258786, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 34935 + }, + { + "epoch": 27.904153354632587, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 34936 + }, + { + "epoch": 27.904952076677315, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 34937 + }, + { + "epoch": 27.905750798722046, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 34938 + }, + { + "epoch": 27.906549520766774, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 34939 + }, + { + "epoch": 27.9073482428115, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 34940 + }, + { + "epoch": 27.90814696485623, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 34941 + }, + { + "epoch": 27.908945686900957, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 34942 + }, + { + "epoch": 27.909744408945688, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 34943 + }, + { + "epoch": 27.910543130990416, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 34944 + }, + { + "epoch": 27.911341853035143, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 34945 + }, + { + "epoch": 27.91214057507987, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 34946 + }, + { + "epoch": 27.912939297124602, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 34947 + }, + { + "epoch": 27.91373801916933, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 34948 + }, + { + "epoch": 27.914536741214057, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 34949 + }, + { + "epoch": 27.915335463258785, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 34950 + }, + { + "epoch": 27.916134185303516, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 34951 + }, + { + "epoch": 27.916932907348244, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 34952 + }, + { + "epoch": 27.91773162939297, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 34953 + }, + { + "epoch": 27.9185303514377, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 34954 + }, + { + "epoch": 27.919329073482427, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 34955 + }, + { + "epoch": 27.920127795527158, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 34956 + }, + { + "epoch": 27.920926517571885, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 34957 + }, + { + "epoch": 27.921725239616613, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 34958 + }, + { + "epoch": 27.92252396166134, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 34959 + }, + { + "epoch": 27.923322683706072, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 34960 + }, + { + "epoch": 27.9241214057508, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0748, + "step": 34961 + }, + { + "epoch": 27.924920127795527, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 34962 + }, + { + "epoch": 27.925718849840255, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 34963 + }, + { + "epoch": 27.926517571884983, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 34964 + }, + { + "epoch": 27.927316293929714, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 34965 + }, + { + "epoch": 27.92811501597444, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 34966 + }, + { + "epoch": 27.92891373801917, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 34967 + }, + { + "epoch": 27.929712460063897, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 34968 + }, + { + "epoch": 27.930511182108628, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 34969 + }, + { + "epoch": 27.931309904153355, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 34970 + }, + { + "epoch": 27.932108626198083, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 34971 + }, + { + "epoch": 27.93290734824281, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 34972 + }, + { + "epoch": 27.93370607028754, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 34973 + }, + { + "epoch": 27.93450479233227, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 34974 + }, + { + "epoch": 27.935303514376997, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 34975 + }, + { + "epoch": 27.936102236421725, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 34976 + }, + { + "epoch": 27.936900958466452, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 34977 + }, + { + "epoch": 27.937699680511184, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 34978 + }, + { + "epoch": 27.93849840255591, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 34979 + }, + { + "epoch": 27.93929712460064, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 34980 + }, + { + "epoch": 27.940095846645367, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0721, + "step": 34981 + }, + { + "epoch": 27.940894568690094, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 34982 + }, + { + "epoch": 27.941693290734825, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 34983 + }, + { + "epoch": 27.942492012779553, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 34984 + }, + { + "epoch": 27.94329073482428, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 34985 + }, + { + "epoch": 27.94408945686901, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 34986 + }, + { + "epoch": 27.94488817891374, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 34987 + }, + { + "epoch": 27.945686900958467, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 34988 + }, + { + "epoch": 27.946485623003195, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 34989 + }, + { + "epoch": 27.947284345047922, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 34990 + }, + { + "epoch": 27.94808306709265, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 34991 + }, + { + "epoch": 27.94888178913738, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 34992 + }, + { + "epoch": 27.94968051118211, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 34993 + }, + { + "epoch": 27.950479233226837, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 34994 + }, + { + "epoch": 27.951277955271564, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 34995 + }, + { + "epoch": 27.952076677316295, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 34996 + }, + { + "epoch": 27.952875399361023, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 34997 + }, + { + "epoch": 27.95367412140575, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 34998 + }, + { + "epoch": 27.95447284345048, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 34999 + }, + { + "epoch": 27.955271565495206, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 35000 + }, + { + "epoch": 27.956070287539937, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 35001 + }, + { + "epoch": 27.956869009584665, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 35002 + }, + { + "epoch": 27.957667731629392, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 35003 + }, + { + "epoch": 27.95846645367412, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 35004 + }, + { + "epoch": 27.95926517571885, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 35005 + }, + { + "epoch": 27.96006389776358, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 35006 + }, + { + "epoch": 27.960862619808307, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 35007 + }, + { + "epoch": 27.961661341853034, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0756, + "step": 35008 + }, + { + "epoch": 27.962460063897765, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 35009 + }, + { + "epoch": 27.963258785942493, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 35010 + }, + { + "epoch": 27.96405750798722, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 35011 + }, + { + "epoch": 27.96485623003195, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 35012 + }, + { + "epoch": 27.965654952076676, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 35013 + }, + { + "epoch": 27.966453674121407, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 35014 + }, + { + "epoch": 27.967252396166135, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 35015 + }, + { + "epoch": 27.968051118210862, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 35016 + }, + { + "epoch": 27.96884984025559, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 35017 + }, + { + "epoch": 27.96964856230032, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 35018 + }, + { + "epoch": 27.97044728434505, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 35019 + }, + { + "epoch": 27.971246006389777, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 35020 + }, + { + "epoch": 27.972044728434504, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35021 + }, + { + "epoch": 27.972843450479232, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 35022 + }, + { + "epoch": 27.973642172523963, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 35023 + }, + { + "epoch": 27.97444089456869, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 35024 + }, + { + "epoch": 27.97523961661342, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 35025 + }, + { + "epoch": 27.976038338658146, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 35026 + }, + { + "epoch": 27.976837060702877, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 35027 + }, + { + "epoch": 27.977635782747605, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 35028 + }, + { + "epoch": 27.978434504792332, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 35029 + }, + { + "epoch": 27.97923322683706, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 35030 + }, + { + "epoch": 27.980031948881788, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 35031 + }, + { + "epoch": 27.98083067092652, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 35032 + }, + { + "epoch": 27.981629392971247, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 35033 + }, + { + "epoch": 27.982428115015974, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 35034 + }, + { + "epoch": 27.983226837060702, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 35035 + }, + { + "epoch": 27.984025559105433, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 35036 + }, + { + "epoch": 27.98482428115016, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 35037 + }, + { + "epoch": 27.98562300319489, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 35038 + }, + { + "epoch": 27.986421725239616, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 35039 + }, + { + "epoch": 27.987220447284344, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 35040 + }, + { + "epoch": 27.988019169329075, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 35041 + }, + { + "epoch": 27.988817891373802, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 35042 + }, + { + "epoch": 27.98961661341853, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 35043 + }, + { + "epoch": 27.990415335463258, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 35044 + }, + { + "epoch": 27.99121405750799, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 35045 + }, + { + "epoch": 27.992012779552716, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 35046 + }, + { + "epoch": 27.992811501597444, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 35047 + }, + { + "epoch": 27.99361022364217, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 35048 + }, + { + "epoch": 27.9944089456869, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 35049 + }, + { + "epoch": 27.99520766773163, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 35050 + }, + { + "epoch": 27.99600638977636, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 35051 + }, + { + "epoch": 27.996805111821086, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 35052 + }, + { + "epoch": 27.997603833865814, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 35053 + }, + { + "epoch": 27.998402555910545, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 35054 + }, + { + "epoch": 27.999201277955272, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35055 + }, + { + "epoch": 28.0, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 35056 + }, + { + "epoch": 28.000798722044728, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 35057 + }, + { + "epoch": 28.001597444089455, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 35058 + }, + { + "epoch": 28.002396166134186, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 35059 + }, + { + "epoch": 28.003194888178914, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 35060 + }, + { + "epoch": 28.00399361022364, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 35061 + }, + { + "epoch": 28.00479233226837, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 35062 + }, + { + "epoch": 28.0055910543131, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 35063 + }, + { + "epoch": 28.00638977635783, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 35064 + }, + { + "epoch": 28.007188498402556, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 35065 + }, + { + "epoch": 28.007987220447284, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 35066 + }, + { + "epoch": 28.00878594249201, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 35067 + }, + { + "epoch": 28.009584664536742, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35068 + }, + { + "epoch": 28.01038338658147, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 35069 + }, + { + "epoch": 28.011182108626198, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 35070 + }, + { + "epoch": 28.011980830670925, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 35071 + }, + { + "epoch": 28.012779552715656, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 35072 + }, + { + "epoch": 28.013578274760384, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 35073 + }, + { + "epoch": 28.01437699680511, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 35074 + }, + { + "epoch": 28.01517571884984, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 35075 + }, + { + "epoch": 28.015974440894567, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 35076 + }, + { + "epoch": 28.016773162939298, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 35077 + }, + { + "epoch": 28.017571884984026, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 35078 + }, + { + "epoch": 28.018370607028753, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 35079 + }, + { + "epoch": 28.01916932907348, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 35080 + }, + { + "epoch": 28.019968051118212, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 35081 + }, + { + "epoch": 28.02076677316294, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 35082 + }, + { + "epoch": 28.021565495207668, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 35083 + }, + { + "epoch": 28.022364217252395, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 35084 + }, + { + "epoch": 28.023162939297123, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 35085 + }, + { + "epoch": 28.023961661341854, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 35086 + }, + { + "epoch": 28.02476038338658, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 35087 + }, + { + "epoch": 28.02555910543131, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 35088 + }, + { + "epoch": 28.026357827476037, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 35089 + }, + { + "epoch": 28.027156549520768, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 35090 + }, + { + "epoch": 28.027955271565496, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 35091 + }, + { + "epoch": 28.028753993610223, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 35092 + }, + { + "epoch": 28.02955271565495, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 35093 + }, + { + "epoch": 28.03035143769968, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 35094 + }, + { + "epoch": 28.03115015974441, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 35095 + }, + { + "epoch": 28.031948881789138, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 35096 + }, + { + "epoch": 28.032747603833865, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 35097 + }, + { + "epoch": 28.033546325878593, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 35098 + }, + { + "epoch": 28.034345047923324, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 35099 + }, + { + "epoch": 28.03514376996805, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 35100 + }, + { + "epoch": 28.03594249201278, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 35101 + }, + { + "epoch": 28.036741214057507, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 35102 + }, + { + "epoch": 28.037539936102238, + "grad_norm": 0.28125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 35103 + }, + { + "epoch": 28.038338658146966, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 35104 + }, + { + "epoch": 28.039137380191693, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 35105 + }, + { + "epoch": 28.03993610223642, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 35106 + }, + { + "epoch": 28.04073482428115, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 35107 + }, + { + "epoch": 28.04153354632588, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 35108 + }, + { + "epoch": 28.042332268370608, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 35109 + }, + { + "epoch": 28.043130990415335, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 35110 + }, + { + "epoch": 28.043929712460063, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 35111 + }, + { + "epoch": 28.044728434504794, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 35112 + }, + { + "epoch": 28.04552715654952, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 35113 + }, + { + "epoch": 28.04632587859425, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 35114 + }, + { + "epoch": 28.047124600638977, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 35115 + }, + { + "epoch": 28.047923322683705, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 35116 + }, + { + "epoch": 28.048722044728436, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 35117 + }, + { + "epoch": 28.049520766773163, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 35118 + }, + { + "epoch": 28.05031948881789, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 35119 + }, + { + "epoch": 28.05111821086262, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 35120 + }, + { + "epoch": 28.05191693290735, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 35121 + }, + { + "epoch": 28.052715654952078, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35122 + }, + { + "epoch": 28.053514376996805, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 35123 + }, + { + "epoch": 28.054313099041533, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35124 + }, + { + "epoch": 28.05511182108626, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 35125 + }, + { + "epoch": 28.05591054313099, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 35126 + }, + { + "epoch": 28.05670926517572, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 35127 + }, + { + "epoch": 28.057507987220447, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 35128 + }, + { + "epoch": 28.058306709265175, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 35129 + }, + { + "epoch": 28.059105431309906, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 35130 + }, + { + "epoch": 28.059904153354633, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0759, + "step": 35131 + }, + { + "epoch": 28.06070287539936, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 35132 + }, + { + "epoch": 28.06150159744409, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 35133 + }, + { + "epoch": 28.062300319488816, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 35134 + }, + { + "epoch": 28.063099041533548, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 35135 + }, + { + "epoch": 28.063897763578275, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 35136 + }, + { + "epoch": 28.064696485623003, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 35137 + }, + { + "epoch": 28.06549520766773, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 35138 + }, + { + "epoch": 28.06629392971246, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 35139 + }, + { + "epoch": 28.06709265175719, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 35140 + }, + { + "epoch": 28.067891373801917, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 35141 + }, + { + "epoch": 28.068690095846645, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 35142 + }, + { + "epoch": 28.069488817891372, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 35143 + }, + { + "epoch": 28.070287539936103, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 35144 + }, + { + "epoch": 28.07108626198083, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 35145 + }, + { + "epoch": 28.07188498402556, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 35146 + }, + { + "epoch": 28.072683706070286, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 35147 + }, + { + "epoch": 28.073482428115017, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 35148 + }, + { + "epoch": 28.074281150159745, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 35149 + }, + { + "epoch": 28.075079872204473, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 35150 + }, + { + "epoch": 28.0758785942492, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 35151 + }, + { + "epoch": 28.076677316293928, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35152 + }, + { + "epoch": 28.07747603833866, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 35153 + }, + { + "epoch": 28.078274760383387, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 35154 + }, + { + "epoch": 28.079073482428115, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 35155 + }, + { + "epoch": 28.079872204472842, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 35156 + }, + { + "epoch": 28.080670926517573, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 35157 + }, + { + "epoch": 28.0814696485623, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 35158 + }, + { + "epoch": 28.08226837060703, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 35159 + }, + { + "epoch": 28.083067092651756, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 35160 + }, + { + "epoch": 28.083865814696484, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 35161 + }, + { + "epoch": 28.084664536741215, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 35162 + }, + { + "epoch": 28.085463258785943, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 35163 + }, + { + "epoch": 28.08626198083067, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 35164 + }, + { + "epoch": 28.087060702875398, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 35165 + }, + { + "epoch": 28.08785942492013, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 35166 + }, + { + "epoch": 28.088658146964857, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 35167 + }, + { + "epoch": 28.089456869009584, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 35168 + }, + { + "epoch": 28.090255591054312, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 35169 + }, + { + "epoch": 28.09105431309904, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 35170 + }, + { + "epoch": 28.09185303514377, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 35171 + }, + { + "epoch": 28.0926517571885, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 35172 + }, + { + "epoch": 28.093450479233226, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 35173 + }, + { + "epoch": 28.094249201277954, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 35174 + }, + { + "epoch": 28.095047923322685, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 35175 + }, + { + "epoch": 28.095846645367413, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 35176 + }, + { + "epoch": 28.09664536741214, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 35177 + }, + { + "epoch": 28.097444089456868, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 35178 + }, + { + "epoch": 28.0982428115016, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35179 + }, + { + "epoch": 28.099041533546327, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 35180 + }, + { + "epoch": 28.099840255591054, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 35181 + }, + { + "epoch": 28.100638977635782, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 35182 + }, + { + "epoch": 28.10143769968051, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35183 + }, + { + "epoch": 28.10223642172524, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 35184 + }, + { + "epoch": 28.10303514376997, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 35185 + }, + { + "epoch": 28.103833865814696, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 35186 + }, + { + "epoch": 28.104632587859424, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.076, + "step": 35187 + }, + { + "epoch": 28.105431309904155, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 35188 + }, + { + "epoch": 28.106230031948883, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 35189 + }, + { + "epoch": 28.10702875399361, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 35190 + }, + { + "epoch": 28.107827476038338, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 35191 + }, + { + "epoch": 28.108626198083066, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 35192 + }, + { + "epoch": 28.109424920127797, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.1077, + "step": 35193 + }, + { + "epoch": 28.110223642172524, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 35194 + }, + { + "epoch": 28.111022364217252, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35195 + }, + { + "epoch": 28.11182108626198, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 35196 + }, + { + "epoch": 28.11261980830671, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 35197 + }, + { + "epoch": 28.11341853035144, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 35198 + }, + { + "epoch": 28.114217252396166, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35199 + }, + { + "epoch": 28.115015974440894, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 35200 + }, + { + "epoch": 28.11581469648562, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 35201 + }, + { + "epoch": 28.116613418530353, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 35202 + }, + { + "epoch": 28.11741214057508, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 35203 + }, + { + "epoch": 28.118210862619808, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 35204 + }, + { + "epoch": 28.119009584664536, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 35205 + }, + { + "epoch": 28.119808306709267, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 35206 + }, + { + "epoch": 28.120607028753994, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 35207 + }, + { + "epoch": 28.121405750798722, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 35208 + }, + { + "epoch": 28.12220447284345, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35209 + }, + { + "epoch": 28.123003194888177, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 35210 + }, + { + "epoch": 28.12380191693291, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 35211 + }, + { + "epoch": 28.124600638977636, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 35212 + }, + { + "epoch": 28.125399361022364, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 35213 + }, + { + "epoch": 28.12619808306709, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 35214 + }, + { + "epoch": 28.126996805111823, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 35215 + }, + { + "epoch": 28.12779552715655, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 35216 + }, + { + "epoch": 28.128594249201278, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 35217 + }, + { + "epoch": 28.129392971246006, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 35218 + }, + { + "epoch": 28.130191693290733, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 35219 + }, + { + "epoch": 28.130990415335464, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 35220 + }, + { + "epoch": 28.131789137380192, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 35221 + }, + { + "epoch": 28.13258785942492, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 35222 + }, + { + "epoch": 28.133386581469647, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35223 + }, + { + "epoch": 28.13418530351438, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 35224 + }, + { + "epoch": 28.134984025559106, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 35225 + }, + { + "epoch": 28.135782747603834, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 35226 + }, + { + "epoch": 28.13658146964856, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 35227 + }, + { + "epoch": 28.13738019169329, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 35228 + }, + { + "epoch": 28.13817891373802, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 35229 + }, + { + "epoch": 28.138977635782748, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 35230 + }, + { + "epoch": 28.139776357827476, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 35231 + }, + { + "epoch": 28.140575079872203, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 35232 + }, + { + "epoch": 28.141373801916934, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 35233 + }, + { + "epoch": 28.142172523961662, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 35234 + }, + { + "epoch": 28.14297124600639, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35235 + }, + { + "epoch": 28.143769968051117, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 35236 + }, + { + "epoch": 28.144568690095845, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 35237 + }, + { + "epoch": 28.145367412140576, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35238 + }, + { + "epoch": 28.146166134185304, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 35239 + }, + { + "epoch": 28.14696485623003, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 35240 + }, + { + "epoch": 28.14776357827476, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 35241 + }, + { + "epoch": 28.14856230031949, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 35242 + }, + { + "epoch": 28.149361022364218, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 35243 + }, + { + "epoch": 28.150159744408946, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 35244 + }, + { + "epoch": 28.150958466453673, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35245 + }, + { + "epoch": 28.1517571884984, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 35246 + }, + { + "epoch": 28.152555910543132, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 35247 + }, + { + "epoch": 28.15335463258786, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 35248 + }, + { + "epoch": 28.154153354632587, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 35249 + }, + { + "epoch": 28.154952076677315, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 35250 + }, + { + "epoch": 28.155750798722046, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 35251 + }, + { + "epoch": 28.156549520766774, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 35252 + }, + { + "epoch": 28.1573482428115, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 35253 + }, + { + "epoch": 28.15814696485623, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 35254 + }, + { + "epoch": 28.15894568690096, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 35255 + }, + { + "epoch": 28.159744408945688, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 35256 + }, + { + "epoch": 28.160543130990416, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 35257 + }, + { + "epoch": 28.161341853035143, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 35258 + }, + { + "epoch": 28.16214057507987, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 35259 + }, + { + "epoch": 28.162939297124602, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 35260 + }, + { + "epoch": 28.16373801916933, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 35261 + }, + { + "epoch": 28.164536741214057, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 35262 + }, + { + "epoch": 28.165335463258785, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0703, + "step": 35263 + }, + { + "epoch": 28.166134185303516, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 35264 + }, + { + "epoch": 28.166932907348244, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 35265 + }, + { + "epoch": 28.16773162939297, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0735, + "step": 35266 + }, + { + "epoch": 28.1685303514377, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 35267 + }, + { + "epoch": 28.169329073482427, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 35268 + }, + { + "epoch": 28.170127795527158, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 35269 + }, + { + "epoch": 28.170926517571885, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 35270 + }, + { + "epoch": 28.171725239616613, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 35271 + }, + { + "epoch": 28.17252396166134, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 35272 + }, + { + "epoch": 28.173322683706072, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 35273 + }, + { + "epoch": 28.1741214057508, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 35274 + }, + { + "epoch": 28.174920127795527, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 35275 + }, + { + "epoch": 28.175718849840255, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 35276 + }, + { + "epoch": 28.176517571884983, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 35277 + }, + { + "epoch": 28.177316293929714, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 35278 + }, + { + "epoch": 28.17811501597444, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 35279 + }, + { + "epoch": 28.17891373801917, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 35280 + }, + { + "epoch": 28.179712460063897, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 35281 + }, + { + "epoch": 28.180511182108628, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 35282 + }, + { + "epoch": 28.181309904153355, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 35283 + }, + { + "epoch": 28.182108626198083, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 35284 + }, + { + "epoch": 28.18290734824281, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 35285 + }, + { + "epoch": 28.18370607028754, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 35286 + }, + { + "epoch": 28.18450479233227, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 35287 + }, + { + "epoch": 28.185303514376997, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 35288 + }, + { + "epoch": 28.186102236421725, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 35289 + }, + { + "epoch": 28.186900958466452, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 35290 + }, + { + "epoch": 28.187699680511184, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 35291 + }, + { + "epoch": 28.18849840255591, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 35292 + }, + { + "epoch": 28.18929712460064, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 35293 + }, + { + "epoch": 28.190095846645367, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 35294 + }, + { + "epoch": 28.190894568690094, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 35295 + }, + { + "epoch": 28.191693290734825, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 35296 + }, + { + "epoch": 28.192492012779553, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 35297 + }, + { + "epoch": 28.19329073482428, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 35298 + }, + { + "epoch": 28.19408945686901, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 35299 + }, + { + "epoch": 28.19488817891374, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 35300 + }, + { + "epoch": 28.195686900958467, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 35301 + }, + { + "epoch": 28.196485623003195, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 35302 + }, + { + "epoch": 28.197284345047922, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 35303 + }, + { + "epoch": 28.19808306709265, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 35304 + }, + { + "epoch": 28.19888178913738, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 35305 + }, + { + "epoch": 28.19968051118211, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 35306 + }, + { + "epoch": 28.200479233226837, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 35307 + }, + { + "epoch": 28.201277955271564, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35308 + }, + { + "epoch": 28.202076677316295, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 35309 + }, + { + "epoch": 28.202875399361023, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 35310 + }, + { + "epoch": 28.20367412140575, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 35311 + }, + { + "epoch": 28.20447284345048, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 35312 + }, + { + "epoch": 28.205271565495206, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 35313 + }, + { + "epoch": 28.206070287539937, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 35314 + }, + { + "epoch": 28.206869009584665, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 35315 + }, + { + "epoch": 28.207667731629392, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 35316 + }, + { + "epoch": 28.20846645367412, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35317 + }, + { + "epoch": 28.20926517571885, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 35318 + }, + { + "epoch": 28.21006389776358, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 35319 + }, + { + "epoch": 28.210862619808307, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 35320 + }, + { + "epoch": 28.211661341853034, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 35321 + }, + { + "epoch": 28.212460063897762, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 35322 + }, + { + "epoch": 28.213258785942493, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 35323 + }, + { + "epoch": 28.21405750798722, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 35324 + }, + { + "epoch": 28.21485623003195, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35325 + }, + { + "epoch": 28.215654952076676, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35326 + }, + { + "epoch": 28.216453674121407, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 35327 + }, + { + "epoch": 28.217252396166135, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 35328 + }, + { + "epoch": 28.218051118210862, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 35329 + }, + { + "epoch": 28.21884984025559, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 35330 + }, + { + "epoch": 28.21964856230032, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 35331 + }, + { + "epoch": 28.22044728434505, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 35332 + }, + { + "epoch": 28.221246006389777, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 35333 + }, + { + "epoch": 28.222044728434504, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 35334 + }, + { + "epoch": 28.222843450479232, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 35335 + }, + { + "epoch": 28.223642172523963, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 35336 + }, + { + "epoch": 28.22444089456869, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 35337 + }, + { + "epoch": 28.22523961661342, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 35338 + }, + { + "epoch": 28.226038338658146, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 35339 + }, + { + "epoch": 28.226837060702877, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 35340 + }, + { + "epoch": 28.227635782747605, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 35341 + }, + { + "epoch": 28.228434504792332, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 35342 + }, + { + "epoch": 28.22923322683706, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 35343 + }, + { + "epoch": 28.230031948881788, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 35344 + }, + { + "epoch": 28.23083067092652, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 35345 + }, + { + "epoch": 28.231629392971247, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 35346 + }, + { + "epoch": 28.232428115015974, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 35347 + }, + { + "epoch": 28.233226837060702, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 35348 + }, + { + "epoch": 28.234025559105433, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 35349 + }, + { + "epoch": 28.23482428115016, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 35350 + }, + { + "epoch": 28.23562300319489, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 35351 + }, + { + "epoch": 28.236421725239616, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 35352 + }, + { + "epoch": 28.237220447284344, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 35353 + }, + { + "epoch": 28.238019169329075, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 35354 + }, + { + "epoch": 28.238817891373802, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 35355 + }, + { + "epoch": 28.23961661341853, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 35356 + }, + { + "epoch": 28.240415335463258, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 35357 + }, + { + "epoch": 28.24121405750799, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 35358 + }, + { + "epoch": 28.242012779552716, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0767, + "step": 35359 + }, + { + "epoch": 28.242811501597444, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 35360 + }, + { + "epoch": 28.24361022364217, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 35361 + }, + { + "epoch": 28.2444089456869, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 35362 + }, + { + "epoch": 28.24520766773163, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 35363 + }, + { + "epoch": 28.24600638977636, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 35364 + }, + { + "epoch": 28.246805111821086, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 35365 + }, + { + "epoch": 28.247603833865814, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 35366 + }, + { + "epoch": 28.248402555910545, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 35367 + }, + { + "epoch": 28.249201277955272, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 35368 + }, + { + "epoch": 28.25, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 35369 + }, + { + "epoch": 28.250798722044728, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 35370 + }, + { + "epoch": 28.251597444089455, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 35371 + }, + { + "epoch": 28.252396166134186, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 35372 + }, + { + "epoch": 28.253194888178914, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 35373 + }, + { + "epoch": 28.25399361022364, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 35374 + }, + { + "epoch": 28.25479233226837, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35375 + }, + { + "epoch": 28.2555910543131, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 35376 + }, + { + "epoch": 28.25638977635783, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35377 + }, + { + "epoch": 28.257188498402556, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 35378 + }, + { + "epoch": 28.257987220447284, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 35379 + }, + { + "epoch": 28.25878594249201, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 35380 + }, + { + "epoch": 28.259584664536742, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 35381 + }, + { + "epoch": 28.26038338658147, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 35382 + }, + { + "epoch": 28.261182108626198, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 35383 + }, + { + "epoch": 28.261980830670925, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 35384 + }, + { + "epoch": 28.262779552715656, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 35385 + }, + { + "epoch": 28.263578274760384, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35386 + }, + { + "epoch": 28.26437699680511, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 35387 + }, + { + "epoch": 28.26517571884984, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 35388 + }, + { + "epoch": 28.265974440894567, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 35389 + }, + { + "epoch": 28.266773162939298, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 35390 + }, + { + "epoch": 28.267571884984026, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 35391 + }, + { + "epoch": 28.268370607028753, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 35392 + }, + { + "epoch": 28.26916932907348, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 35393 + }, + { + "epoch": 28.269968051118212, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 35394 + }, + { + "epoch": 28.27076677316294, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 35395 + }, + { + "epoch": 28.271565495207668, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 35396 + }, + { + "epoch": 28.272364217252395, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 35397 + }, + { + "epoch": 28.273162939297123, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 35398 + }, + { + "epoch": 28.273961661341854, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 35399 + }, + { + "epoch": 28.27476038338658, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35400 + }, + { + "epoch": 28.27555910543131, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 35401 + }, + { + "epoch": 28.276357827476037, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 35402 + }, + { + "epoch": 28.277156549520768, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 35403 + }, + { + "epoch": 28.277955271565496, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 35404 + }, + { + "epoch": 28.278753993610223, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 35405 + }, + { + "epoch": 28.27955271565495, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 35406 + }, + { + "epoch": 28.28035143769968, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 35407 + }, + { + "epoch": 28.28115015974441, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 35408 + }, + { + "epoch": 28.281948881789138, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 35409 + }, + { + "epoch": 28.282747603833865, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 35410 + }, + { + "epoch": 28.283546325878593, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 35411 + }, + { + "epoch": 28.284345047923324, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 35412 + }, + { + "epoch": 28.28514376996805, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35413 + }, + { + "epoch": 28.28594249201278, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 35414 + }, + { + "epoch": 28.286741214057507, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 35415 + }, + { + "epoch": 28.287539936102238, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 35416 + }, + { + "epoch": 28.288338658146966, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 35417 + }, + { + "epoch": 28.289137380191693, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 35418 + }, + { + "epoch": 28.28993610223642, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35419 + }, + { + "epoch": 28.29073482428115, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 35420 + }, + { + "epoch": 28.29153354632588, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 35421 + }, + { + "epoch": 28.292332268370608, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.104, + "step": 35422 + }, + { + "epoch": 28.293130990415335, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 35423 + }, + { + "epoch": 28.293929712460063, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 35424 + }, + { + "epoch": 28.294728434504794, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 35425 + }, + { + "epoch": 28.29552715654952, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 35426 + }, + { + "epoch": 28.29632587859425, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 35427 + }, + { + "epoch": 28.297124600638977, + "grad_norm": 0.09130859375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 35428 + }, + { + "epoch": 28.297923322683705, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35429 + }, + { + "epoch": 28.298722044728436, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35430 + }, + { + "epoch": 28.299520766773163, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 35431 + }, + { + "epoch": 28.30031948881789, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 35432 + }, + { + "epoch": 28.30111821086262, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35433 + }, + { + "epoch": 28.30191693290735, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 35434 + }, + { + "epoch": 28.302715654952078, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 35435 + }, + { + "epoch": 28.303514376996805, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 35436 + }, + { + "epoch": 28.304313099041533, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 35437 + }, + { + "epoch": 28.30511182108626, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 35438 + }, + { + "epoch": 28.30591054313099, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 35439 + }, + { + "epoch": 28.30670926517572, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 35440 + }, + { + "epoch": 28.307507987220447, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 35441 + }, + { + "epoch": 28.308306709265175, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0724, + "step": 35442 + }, + { + "epoch": 28.309105431309906, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 35443 + }, + { + "epoch": 28.309904153354633, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 35444 + }, + { + "epoch": 28.31070287539936, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 35445 + }, + { + "epoch": 28.31150159744409, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 35446 + }, + { + "epoch": 28.312300319488816, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 35447 + }, + { + "epoch": 28.313099041533548, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 35448 + }, + { + "epoch": 28.313897763578275, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 35449 + }, + { + "epoch": 28.314696485623003, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 35450 + }, + { + "epoch": 28.31549520766773, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0763, + "step": 35451 + }, + { + "epoch": 28.31629392971246, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 35452 + }, + { + "epoch": 28.31709265175719, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 35453 + }, + { + "epoch": 28.317891373801917, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 35454 + }, + { + "epoch": 28.318690095846645, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 35455 + }, + { + "epoch": 28.319488817891372, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 35456 + }, + { + "epoch": 28.320287539936103, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 35457 + }, + { + "epoch": 28.32108626198083, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 35458 + }, + { + "epoch": 28.32188498402556, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 35459 + }, + { + "epoch": 28.322683706070286, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0747, + "step": 35460 + }, + { + "epoch": 28.323482428115017, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 35461 + }, + { + "epoch": 28.324281150159745, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 35462 + }, + { + "epoch": 28.325079872204473, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 35463 + }, + { + "epoch": 28.3258785942492, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 35464 + }, + { + "epoch": 28.326677316293928, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 35465 + }, + { + "epoch": 28.32747603833866, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 35466 + }, + { + "epoch": 28.328274760383387, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 35467 + }, + { + "epoch": 28.329073482428115, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 35468 + }, + { + "epoch": 28.329872204472842, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 35469 + }, + { + "epoch": 28.330670926517573, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 35470 + }, + { + "epoch": 28.3314696485623, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 35471 + }, + { + "epoch": 28.33226837060703, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 35472 + }, + { + "epoch": 28.333067092651756, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 35473 + }, + { + "epoch": 28.333865814696484, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 35474 + }, + { + "epoch": 28.334664536741215, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 35475 + }, + { + "epoch": 28.335463258785943, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 35476 + }, + { + "epoch": 28.33626198083067, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 35477 + }, + { + "epoch": 28.337060702875398, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 35478 + }, + { + "epoch": 28.33785942492013, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 35479 + }, + { + "epoch": 28.338658146964857, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 35480 + }, + { + "epoch": 28.339456869009584, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35481 + }, + { + "epoch": 28.340255591054312, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 35482 + }, + { + "epoch": 28.341054313099043, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 35483 + }, + { + "epoch": 28.34185303514377, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 35484 + }, + { + "epoch": 28.3426517571885, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 35485 + }, + { + "epoch": 28.343450479233226, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 35486 + }, + { + "epoch": 28.344249201277954, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 35487 + }, + { + "epoch": 28.345047923322685, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 35488 + }, + { + "epoch": 28.345846645367413, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 35489 + }, + { + "epoch": 28.34664536741214, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 35490 + }, + { + "epoch": 28.347444089456868, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 35491 + }, + { + "epoch": 28.3482428115016, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 35492 + }, + { + "epoch": 28.349041533546327, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 35493 + }, + { + "epoch": 28.349840255591054, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 35494 + }, + { + "epoch": 28.350638977635782, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 35495 + }, + { + "epoch": 28.35143769968051, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 35496 + }, + { + "epoch": 28.35223642172524, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 35497 + }, + { + "epoch": 28.35303514376997, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 35498 + }, + { + "epoch": 28.353833865814696, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 35499 + }, + { + "epoch": 28.354632587859424, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 35500 + }, + { + "epoch": 28.355431309904155, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 35501 + }, + { + "epoch": 28.356230031948883, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 35502 + }, + { + "epoch": 28.35702875399361, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 35503 + }, + { + "epoch": 28.357827476038338, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 35504 + }, + { + "epoch": 28.358626198083066, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 35505 + }, + { + "epoch": 28.359424920127797, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 35506 + }, + { + "epoch": 28.360223642172524, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 35507 + }, + { + "epoch": 28.361022364217252, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 35508 + }, + { + "epoch": 28.36182108626198, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 35509 + }, + { + "epoch": 28.36261980830671, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 35510 + }, + { + "epoch": 28.36341853035144, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 35511 + }, + { + "epoch": 28.364217252396166, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 35512 + }, + { + "epoch": 28.365015974440894, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 35513 + }, + { + "epoch": 28.36581469648562, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 35514 + }, + { + "epoch": 28.366613418530353, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 35515 + }, + { + "epoch": 28.36741214057508, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 35516 + }, + { + "epoch": 28.368210862619808, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 35517 + }, + { + "epoch": 28.369009584664536, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 35518 + }, + { + "epoch": 28.369808306709267, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 35519 + }, + { + "epoch": 28.370607028753994, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 35520 + }, + { + "epoch": 28.371405750798722, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0748, + "step": 35521 + }, + { + "epoch": 28.37220447284345, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 35522 + }, + { + "epoch": 28.373003194888177, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 35523 + }, + { + "epoch": 28.37380191693291, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 35524 + }, + { + "epoch": 28.374600638977636, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 35525 + }, + { + "epoch": 28.375399361022364, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35526 + }, + { + "epoch": 28.37619808306709, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 35527 + }, + { + "epoch": 28.376996805111823, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 35528 + }, + { + "epoch": 28.37779552715655, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 35529 + }, + { + "epoch": 28.378594249201278, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 35530 + }, + { + "epoch": 28.379392971246006, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 35531 + }, + { + "epoch": 28.380191693290733, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 35532 + }, + { + "epoch": 28.380990415335464, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35533 + }, + { + "epoch": 28.381789137380192, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 35534 + }, + { + "epoch": 28.38258785942492, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 35535 + }, + { + "epoch": 28.383386581469647, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 35536 + }, + { + "epoch": 28.38418530351438, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 35537 + }, + { + "epoch": 28.384984025559106, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 35538 + }, + { + "epoch": 28.385782747603834, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35539 + }, + { + "epoch": 28.38658146964856, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 35540 + }, + { + "epoch": 28.38738019169329, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35541 + }, + { + "epoch": 28.38817891373802, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 35542 + }, + { + "epoch": 28.388977635782748, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 35543 + }, + { + "epoch": 28.389776357827476, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 35544 + }, + { + "epoch": 28.390575079872203, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 35545 + }, + { + "epoch": 28.391373801916934, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 35546 + }, + { + "epoch": 28.392172523961662, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35547 + }, + { + "epoch": 28.39297124600639, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 35548 + }, + { + "epoch": 28.393769968051117, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 35549 + }, + { + "epoch": 28.394568690095845, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 35550 + }, + { + "epoch": 28.395367412140576, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 35551 + }, + { + "epoch": 28.396166134185304, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35552 + }, + { + "epoch": 28.39696485623003, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 35553 + }, + { + "epoch": 28.39776357827476, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 35554 + }, + { + "epoch": 28.39856230031949, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 35555 + }, + { + "epoch": 28.399361022364218, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 35556 + }, + { + "epoch": 28.400159744408946, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 35557 + }, + { + "epoch": 28.400958466453673, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 35558 + }, + { + "epoch": 28.4017571884984, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 35559 + }, + { + "epoch": 28.402555910543132, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 35560 + }, + { + "epoch": 28.40335463258786, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 35561 + }, + { + "epoch": 28.404153354632587, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 35562 + }, + { + "epoch": 28.404952076677315, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 35563 + }, + { + "epoch": 28.405750798722046, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 35564 + }, + { + "epoch": 28.406549520766774, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 35565 + }, + { + "epoch": 28.4073482428115, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 35566 + }, + { + "epoch": 28.40814696485623, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 35567 + }, + { + "epoch": 28.408945686900957, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 35568 + }, + { + "epoch": 28.409744408945688, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 35569 + }, + { + "epoch": 28.410543130990416, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 35570 + }, + { + "epoch": 28.411341853035143, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35571 + }, + { + "epoch": 28.41214057507987, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 35572 + }, + { + "epoch": 28.412939297124602, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 35573 + }, + { + "epoch": 28.41373801916933, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 35574 + }, + { + "epoch": 28.414536741214057, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 35575 + }, + { + "epoch": 28.415335463258785, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 35576 + }, + { + "epoch": 28.416134185303516, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 35577 + }, + { + "epoch": 28.416932907348244, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 35578 + }, + { + "epoch": 28.41773162939297, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 35579 + }, + { + "epoch": 28.4185303514377, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 35580 + }, + { + "epoch": 28.419329073482427, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35581 + }, + { + "epoch": 28.420127795527158, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 35582 + }, + { + "epoch": 28.420926517571885, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 35583 + }, + { + "epoch": 28.421725239616613, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 35584 + }, + { + "epoch": 28.42252396166134, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 35585 + }, + { + "epoch": 28.423322683706072, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 35586 + }, + { + "epoch": 28.4241214057508, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 35587 + }, + { + "epoch": 28.424920127795527, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 35588 + }, + { + "epoch": 28.425718849840255, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 35589 + }, + { + "epoch": 28.426517571884983, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 35590 + }, + { + "epoch": 28.427316293929714, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 35591 + }, + { + "epoch": 28.42811501597444, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 35592 + }, + { + "epoch": 28.42891373801917, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 35593 + }, + { + "epoch": 28.429712460063897, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 35594 + }, + { + "epoch": 28.430511182108628, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 35595 + }, + { + "epoch": 28.431309904153355, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 35596 + }, + { + "epoch": 28.432108626198083, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 35597 + }, + { + "epoch": 28.43290734824281, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 35598 + }, + { + "epoch": 28.43370607028754, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 35599 + }, + { + "epoch": 28.43450479233227, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 35600 + }, + { + "epoch": 28.435303514376997, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 35601 + }, + { + "epoch": 28.436102236421725, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 35602 + }, + { + "epoch": 28.436900958466452, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 35603 + }, + { + "epoch": 28.437699680511184, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 35604 + }, + { + "epoch": 28.43849840255591, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 35605 + }, + { + "epoch": 28.43929712460064, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 35606 + }, + { + "epoch": 28.440095846645367, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 35607 + }, + { + "epoch": 28.440894568690094, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 35608 + }, + { + "epoch": 28.441693290734825, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 35609 + }, + { + "epoch": 28.442492012779553, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 35610 + }, + { + "epoch": 28.44329073482428, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 35611 + }, + { + "epoch": 28.44408945686901, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 35612 + }, + { + "epoch": 28.44488817891374, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 35613 + }, + { + "epoch": 28.445686900958467, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 35614 + }, + { + "epoch": 28.446485623003195, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 35615 + }, + { + "epoch": 28.447284345047922, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 35616 + }, + { + "epoch": 28.44808306709265, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 35617 + }, + { + "epoch": 28.44888178913738, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 35618 + }, + { + "epoch": 28.44968051118211, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35619 + }, + { + "epoch": 28.450479233226837, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 35620 + }, + { + "epoch": 28.451277955271564, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 35621 + }, + { + "epoch": 28.452076677316295, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 35622 + }, + { + "epoch": 28.452875399361023, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 35623 + }, + { + "epoch": 28.45367412140575, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 35624 + }, + { + "epoch": 28.45447284345048, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 35625 + }, + { + "epoch": 28.455271565495206, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 35626 + }, + { + "epoch": 28.456070287539937, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 35627 + }, + { + "epoch": 28.456869009584665, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 35628 + }, + { + "epoch": 28.457667731629392, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 35629 + }, + { + "epoch": 28.45846645367412, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 35630 + }, + { + "epoch": 28.45926517571885, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 35631 + }, + { + "epoch": 28.46006389776358, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35632 + }, + { + "epoch": 28.460862619808307, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35633 + }, + { + "epoch": 28.461661341853034, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 35634 + }, + { + "epoch": 28.462460063897762, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 35635 + }, + { + "epoch": 28.463258785942493, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 35636 + }, + { + "epoch": 28.46405750798722, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 35637 + }, + { + "epoch": 28.46485623003195, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 35638 + }, + { + "epoch": 28.465654952076676, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 35639 + }, + { + "epoch": 28.466453674121407, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 35640 + }, + { + "epoch": 28.467252396166135, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35641 + }, + { + "epoch": 28.468051118210862, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 35642 + }, + { + "epoch": 28.46884984025559, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 35643 + }, + { + "epoch": 28.46964856230032, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.073, + "step": 35644 + }, + { + "epoch": 28.47044728434505, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 35645 + }, + { + "epoch": 28.471246006389777, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 35646 + }, + { + "epoch": 28.472044728434504, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 35647 + }, + { + "epoch": 28.472843450479232, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 35648 + }, + { + "epoch": 28.473642172523963, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 35649 + }, + { + "epoch": 28.47444089456869, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 35650 + }, + { + "epoch": 28.47523961661342, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 35651 + }, + { + "epoch": 28.476038338658146, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 35652 + }, + { + "epoch": 28.476837060702877, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 35653 + }, + { + "epoch": 28.477635782747605, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 35654 + }, + { + "epoch": 28.478434504792332, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 35655 + }, + { + "epoch": 28.47923322683706, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 35656 + }, + { + "epoch": 28.480031948881788, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 35657 + }, + { + "epoch": 28.48083067092652, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35658 + }, + { + "epoch": 28.481629392971247, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 35659 + }, + { + "epoch": 28.482428115015974, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 35660 + }, + { + "epoch": 28.483226837060702, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0761, + "step": 35661 + }, + { + "epoch": 28.484025559105433, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 35662 + }, + { + "epoch": 28.48482428115016, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35663 + }, + { + "epoch": 28.48562300319489, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 35664 + }, + { + "epoch": 28.486421725239616, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 35665 + }, + { + "epoch": 28.487220447284344, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 35666 + }, + { + "epoch": 28.488019169329075, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 35667 + }, + { + "epoch": 28.488817891373802, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 35668 + }, + { + "epoch": 28.48961661341853, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 35669 + }, + { + "epoch": 28.490415335463258, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35670 + }, + { + "epoch": 28.49121405750799, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 35671 + }, + { + "epoch": 28.492012779552716, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 35672 + }, + { + "epoch": 28.492811501597444, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 35673 + }, + { + "epoch": 28.49361022364217, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 35674 + }, + { + "epoch": 28.4944089456869, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0753, + "step": 35675 + }, + { + "epoch": 28.49520766773163, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 35676 + }, + { + "epoch": 28.49600638977636, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 35677 + }, + { + "epoch": 28.496805111821086, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 35678 + }, + { + "epoch": 28.497603833865814, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 35679 + }, + { + "epoch": 28.498402555910545, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 35680 + }, + { + "epoch": 28.499201277955272, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 35681 + }, + { + "epoch": 28.5, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 35682 + }, + { + "epoch": 28.500798722044728, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35683 + }, + { + "epoch": 28.501597444089455, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 35684 + }, + { + "epoch": 28.502396166134186, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35685 + }, + { + "epoch": 28.503194888178914, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 35686 + }, + { + "epoch": 28.50399361022364, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 35687 + }, + { + "epoch": 28.50479233226837, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 35688 + }, + { + "epoch": 28.5055910543131, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 35689 + }, + { + "epoch": 28.50638977635783, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 35690 + }, + { + "epoch": 28.507188498402556, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 35691 + }, + { + "epoch": 28.507987220447284, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 35692 + }, + { + "epoch": 28.50878594249201, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 35693 + }, + { + "epoch": 28.509584664536742, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 35694 + }, + { + "epoch": 28.51038338658147, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 35695 + }, + { + "epoch": 28.511182108626198, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 35696 + }, + { + "epoch": 28.511980830670925, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 35697 + }, + { + "epoch": 28.512779552715656, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 35698 + }, + { + "epoch": 28.513578274760384, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35699 + }, + { + "epoch": 28.51437699680511, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 35700 + }, + { + "epoch": 28.51517571884984, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 35701 + }, + { + "epoch": 28.515974440894567, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 35702 + }, + { + "epoch": 28.516773162939298, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 35703 + }, + { + "epoch": 28.517571884984026, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 35704 + }, + { + "epoch": 28.518370607028753, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 35705 + }, + { + "epoch": 28.51916932907348, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 35706 + }, + { + "epoch": 28.519968051118212, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 35707 + }, + { + "epoch": 28.52076677316294, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 35708 + }, + { + "epoch": 28.521565495207668, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 35709 + }, + { + "epoch": 28.522364217252395, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 35710 + }, + { + "epoch": 28.523162939297123, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 35711 + }, + { + "epoch": 28.523961661341854, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 35712 + }, + { + "epoch": 28.52476038338658, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35713 + }, + { + "epoch": 28.52555910543131, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 35714 + }, + { + "epoch": 28.526357827476037, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 35715 + }, + { + "epoch": 28.527156549520768, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 35716 + }, + { + "epoch": 28.527955271565496, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 35717 + }, + { + "epoch": 28.528753993610223, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 35718 + }, + { + "epoch": 28.52955271565495, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 35719 + }, + { + "epoch": 28.53035143769968, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.1062, + "step": 35720 + }, + { + "epoch": 28.53115015974441, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 35721 + }, + { + "epoch": 28.531948881789138, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 35722 + }, + { + "epoch": 28.532747603833865, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 35723 + }, + { + "epoch": 28.533546325878593, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35724 + }, + { + "epoch": 28.534345047923324, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 35725 + }, + { + "epoch": 28.53514376996805, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 35726 + }, + { + "epoch": 28.53594249201278, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 35727 + }, + { + "epoch": 28.536741214057507, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 35728 + }, + { + "epoch": 28.537539936102235, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 35729 + }, + { + "epoch": 28.538338658146966, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 35730 + }, + { + "epoch": 28.539137380191693, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 35731 + }, + { + "epoch": 28.53993610223642, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 35732 + }, + { + "epoch": 28.54073482428115, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 35733 + }, + { + "epoch": 28.54153354632588, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 35734 + }, + { + "epoch": 28.542332268370608, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 35735 + }, + { + "epoch": 28.543130990415335, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35736 + }, + { + "epoch": 28.543929712460063, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 35737 + }, + { + "epoch": 28.544728434504794, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 35738 + }, + { + "epoch": 28.54552715654952, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35739 + }, + { + "epoch": 28.54632587859425, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 35740 + }, + { + "epoch": 28.547124600638977, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 35741 + }, + { + "epoch": 28.547923322683705, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 35742 + }, + { + "epoch": 28.548722044728436, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 35743 + }, + { + "epoch": 28.549520766773163, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 35744 + }, + { + "epoch": 28.55031948881789, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 35745 + }, + { + "epoch": 28.55111821086262, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35746 + }, + { + "epoch": 28.55191693290735, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 35747 + }, + { + "epoch": 28.552715654952078, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 35748 + }, + { + "epoch": 28.553514376996805, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 35749 + }, + { + "epoch": 28.554313099041533, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 35750 + }, + { + "epoch": 28.55511182108626, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 35751 + }, + { + "epoch": 28.55591054313099, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 35752 + }, + { + "epoch": 28.55670926517572, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 35753 + }, + { + "epoch": 28.557507987220447, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35754 + }, + { + "epoch": 28.558306709265175, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 35755 + }, + { + "epoch": 28.559105431309906, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 35756 + }, + { + "epoch": 28.559904153354633, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 35757 + }, + { + "epoch": 28.56070287539936, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 35758 + }, + { + "epoch": 28.56150159744409, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 35759 + }, + { + "epoch": 28.562300319488816, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35760 + }, + { + "epoch": 28.563099041533548, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 35761 + }, + { + "epoch": 28.563897763578275, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 35762 + }, + { + "epoch": 28.564696485623003, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0756, + "step": 35763 + }, + { + "epoch": 28.56549520766773, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 35764 + }, + { + "epoch": 28.56629392971246, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 35765 + }, + { + "epoch": 28.56709265175719, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 35766 + }, + { + "epoch": 28.567891373801917, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 35767 + }, + { + "epoch": 28.568690095846645, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 35768 + }, + { + "epoch": 28.569488817891372, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 35769 + }, + { + "epoch": 28.570287539936103, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 35770 + }, + { + "epoch": 28.57108626198083, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35771 + }, + { + "epoch": 28.57188498402556, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 35772 + }, + { + "epoch": 28.572683706070286, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 35773 + }, + { + "epoch": 28.573482428115017, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 35774 + }, + { + "epoch": 28.574281150159745, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 35775 + }, + { + "epoch": 28.575079872204473, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 35776 + }, + { + "epoch": 28.5758785942492, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 35777 + }, + { + "epoch": 28.576677316293928, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 35778 + }, + { + "epoch": 28.57747603833866, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 35779 + }, + { + "epoch": 28.578274760383387, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 35780 + }, + { + "epoch": 28.579073482428115, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 35781 + }, + { + "epoch": 28.579872204472842, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 35782 + }, + { + "epoch": 28.580670926517573, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 35783 + }, + { + "epoch": 28.5814696485623, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 35784 + }, + { + "epoch": 28.58226837060703, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 35785 + }, + { + "epoch": 28.583067092651756, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35786 + }, + { + "epoch": 28.583865814696484, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 35787 + }, + { + "epoch": 28.584664536741215, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 35788 + }, + { + "epoch": 28.585463258785943, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 35789 + }, + { + "epoch": 28.58626198083067, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 35790 + }, + { + "epoch": 28.587060702875398, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 35791 + }, + { + "epoch": 28.58785942492013, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 35792 + }, + { + "epoch": 28.588658146964857, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 35793 + }, + { + "epoch": 28.589456869009584, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35794 + }, + { + "epoch": 28.590255591054312, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 35795 + }, + { + "epoch": 28.591054313099043, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 35796 + }, + { + "epoch": 28.59185303514377, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 35797 + }, + { + "epoch": 28.5926517571885, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 35798 + }, + { + "epoch": 28.593450479233226, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 35799 + }, + { + "epoch": 28.594249201277954, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 35800 + }, + { + "epoch": 28.595047923322685, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 35801 + }, + { + "epoch": 28.595846645367413, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 35802 + }, + { + "epoch": 28.59664536741214, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 35803 + }, + { + "epoch": 28.597444089456868, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 35804 + }, + { + "epoch": 28.5982428115016, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 35805 + }, + { + "epoch": 28.599041533546327, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 35806 + }, + { + "epoch": 28.599840255591054, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 35807 + }, + { + "epoch": 28.600638977635782, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 35808 + }, + { + "epoch": 28.60143769968051, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 35809 + }, + { + "epoch": 28.60223642172524, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 35810 + }, + { + "epoch": 28.60303514376997, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 35811 + }, + { + "epoch": 28.603833865814696, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 35812 + }, + { + "epoch": 28.604632587859424, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 35813 + }, + { + "epoch": 28.605431309904155, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 35814 + }, + { + "epoch": 28.606230031948883, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 35815 + }, + { + "epoch": 28.60702875399361, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 35816 + }, + { + "epoch": 28.607827476038338, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 35817 + }, + { + "epoch": 28.608626198083066, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 35818 + }, + { + "epoch": 28.609424920127797, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 35819 + }, + { + "epoch": 28.610223642172524, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 35820 + }, + { + "epoch": 28.611022364217252, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 35821 + }, + { + "epoch": 28.61182108626198, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 35822 + }, + { + "epoch": 28.61261980830671, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 35823 + }, + { + "epoch": 28.61341853035144, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 35824 + }, + { + "epoch": 28.614217252396166, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 35825 + }, + { + "epoch": 28.615015974440894, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 35826 + }, + { + "epoch": 28.61581469648562, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 35827 + }, + { + "epoch": 28.616613418530353, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 35828 + }, + { + "epoch": 28.61741214057508, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 35829 + }, + { + "epoch": 28.618210862619808, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 35830 + }, + { + "epoch": 28.619009584664536, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 35831 + }, + { + "epoch": 28.619808306709267, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 35832 + }, + { + "epoch": 28.620607028753994, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.069, + "step": 35833 + }, + { + "epoch": 28.621405750798722, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 35834 + }, + { + "epoch": 28.62220447284345, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 35835 + }, + { + "epoch": 28.623003194888177, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 35836 + }, + { + "epoch": 28.62380191693291, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 35837 + }, + { + "epoch": 28.624600638977636, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 35838 + }, + { + "epoch": 28.625399361022364, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 35839 + }, + { + "epoch": 28.62619808306709, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 35840 + }, + { + "epoch": 28.626996805111823, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 35841 + }, + { + "epoch": 28.62779552715655, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 35842 + }, + { + "epoch": 28.628594249201278, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 35843 + }, + { + "epoch": 28.629392971246006, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 35844 + }, + { + "epoch": 28.630191693290733, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35845 + }, + { + "epoch": 28.630990415335464, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 35846 + }, + { + "epoch": 28.631789137380192, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 35847 + }, + { + "epoch": 28.63258785942492, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 35848 + }, + { + "epoch": 28.633386581469647, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 35849 + }, + { + "epoch": 28.63418530351438, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 35850 + }, + { + "epoch": 28.634984025559106, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 35851 + }, + { + "epoch": 28.635782747603834, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 35852 + }, + { + "epoch": 28.63658146964856, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 35853 + }, + { + "epoch": 28.63738019169329, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 35854 + }, + { + "epoch": 28.63817891373802, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 35855 + }, + { + "epoch": 28.638977635782748, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 35856 + }, + { + "epoch": 28.639776357827476, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0754, + "step": 35857 + }, + { + "epoch": 28.640575079872203, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 35858 + }, + { + "epoch": 28.641373801916934, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 35859 + }, + { + "epoch": 28.642172523961662, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 35860 + }, + { + "epoch": 28.64297124600639, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 35861 + }, + { + "epoch": 28.643769968051117, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 35862 + }, + { + "epoch": 28.644568690095845, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 35863 + }, + { + "epoch": 28.645367412140576, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 35864 + }, + { + "epoch": 28.646166134185304, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 35865 + }, + { + "epoch": 28.64696485623003, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35866 + }, + { + "epoch": 28.64776357827476, + "grad_norm": 0.0986328125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 35867 + }, + { + "epoch": 28.64856230031949, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 35868 + }, + { + "epoch": 28.649361022364218, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 35869 + }, + { + "epoch": 28.650159744408946, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 35870 + }, + { + "epoch": 28.650958466453673, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 35871 + }, + { + "epoch": 28.6517571884984, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 35872 + }, + { + "epoch": 28.652555910543132, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 35873 + }, + { + "epoch": 28.65335463258786, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 35874 + }, + { + "epoch": 28.654153354632587, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 35875 + }, + { + "epoch": 28.654952076677315, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 35876 + }, + { + "epoch": 28.655750798722046, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 35877 + }, + { + "epoch": 28.656549520766774, + "grad_norm": 0.2353515625, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 35878 + }, + { + "epoch": 28.6573482428115, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 35879 + }, + { + "epoch": 28.65814696485623, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 35880 + }, + { + "epoch": 28.658945686900957, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 35881 + }, + { + "epoch": 28.659744408945688, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 35882 + }, + { + "epoch": 28.660543130990416, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35883 + }, + { + "epoch": 28.661341853035143, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 35884 + }, + { + "epoch": 28.66214057507987, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 35885 + }, + { + "epoch": 28.662939297124602, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 35886 + }, + { + "epoch": 28.66373801916933, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 35887 + }, + { + "epoch": 28.664536741214057, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 35888 + }, + { + "epoch": 28.665335463258785, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35889 + }, + { + "epoch": 28.666134185303516, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 35890 + }, + { + "epoch": 28.666932907348244, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 35891 + }, + { + "epoch": 28.66773162939297, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 35892 + }, + { + "epoch": 28.6685303514377, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 35893 + }, + { + "epoch": 28.669329073482427, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0735, + "step": 35894 + }, + { + "epoch": 28.670127795527158, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 35895 + }, + { + "epoch": 28.670926517571885, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 35896 + }, + { + "epoch": 28.671725239616613, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 35897 + }, + { + "epoch": 28.67252396166134, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 35898 + }, + { + "epoch": 28.673322683706072, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 35899 + }, + { + "epoch": 28.6741214057508, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 35900 + }, + { + "epoch": 28.674920127795527, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 35901 + }, + { + "epoch": 28.675718849840255, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0752, + "step": 35902 + }, + { + "epoch": 28.676517571884983, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 35903 + }, + { + "epoch": 28.677316293929714, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 35904 + }, + { + "epoch": 28.67811501597444, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35905 + }, + { + "epoch": 28.67891373801917, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 35906 + }, + { + "epoch": 28.679712460063897, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35907 + }, + { + "epoch": 28.680511182108628, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 35908 + }, + { + "epoch": 28.681309904153355, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 35909 + }, + { + "epoch": 28.682108626198083, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 35910 + }, + { + "epoch": 28.68290734824281, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 35911 + }, + { + "epoch": 28.68370607028754, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 35912 + }, + { + "epoch": 28.68450479233227, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 35913 + }, + { + "epoch": 28.685303514376997, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 35914 + }, + { + "epoch": 28.686102236421725, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 35915 + }, + { + "epoch": 28.686900958466452, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 35916 + }, + { + "epoch": 28.687699680511184, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 35917 + }, + { + "epoch": 28.68849840255591, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35918 + }, + { + "epoch": 28.68929712460064, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 35919 + }, + { + "epoch": 28.690095846645367, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 35920 + }, + { + "epoch": 28.690894568690094, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 35921 + }, + { + "epoch": 28.691693290734825, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 35922 + }, + { + "epoch": 28.692492012779553, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 35923 + }, + { + "epoch": 28.69329073482428, + "grad_norm": 0.095703125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 35924 + }, + { + "epoch": 28.69408945686901, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 35925 + }, + { + "epoch": 28.69488817891374, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 35926 + }, + { + "epoch": 28.695686900958467, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 35927 + }, + { + "epoch": 28.696485623003195, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 35928 + }, + { + "epoch": 28.697284345047922, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 35929 + }, + { + "epoch": 28.69808306709265, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 35930 + }, + { + "epoch": 28.69888178913738, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 35931 + }, + { + "epoch": 28.69968051118211, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 35932 + }, + { + "epoch": 28.700479233226837, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 35933 + }, + { + "epoch": 28.701277955271564, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 35934 + }, + { + "epoch": 28.702076677316295, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 35935 + }, + { + "epoch": 28.702875399361023, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 35936 + }, + { + "epoch": 28.70367412140575, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 35937 + }, + { + "epoch": 28.70447284345048, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 35938 + }, + { + "epoch": 28.705271565495206, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 35939 + }, + { + "epoch": 28.706070287539937, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 35940 + }, + { + "epoch": 28.706869009584665, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 35941 + }, + { + "epoch": 28.707667731629392, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 35942 + }, + { + "epoch": 28.70846645367412, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 35943 + }, + { + "epoch": 28.70926517571885, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 35944 + }, + { + "epoch": 28.71006389776358, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 35945 + }, + { + "epoch": 28.710862619808307, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 35946 + }, + { + "epoch": 28.711661341853034, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 35947 + }, + { + "epoch": 28.712460063897765, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 35948 + }, + { + "epoch": 28.713258785942493, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 35949 + }, + { + "epoch": 28.71405750798722, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 35950 + }, + { + "epoch": 28.71485623003195, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 35951 + }, + { + "epoch": 28.715654952076676, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 35952 + }, + { + "epoch": 28.716453674121407, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 35953 + }, + { + "epoch": 28.717252396166135, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 35954 + }, + { + "epoch": 28.718051118210862, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 35955 + }, + { + "epoch": 28.71884984025559, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0742, + "step": 35956 + }, + { + "epoch": 28.71964856230032, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 35957 + }, + { + "epoch": 28.72044728434505, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 35958 + }, + { + "epoch": 28.721246006389777, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 35959 + }, + { + "epoch": 28.722044728434504, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 35960 + }, + { + "epoch": 28.722843450479232, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 35961 + }, + { + "epoch": 28.723642172523963, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 35962 + }, + { + "epoch": 28.72444089456869, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 35963 + }, + { + "epoch": 28.72523961661342, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 35964 + }, + { + "epoch": 28.726038338658146, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 35965 + }, + { + "epoch": 28.726837060702877, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 35966 + }, + { + "epoch": 28.727635782747605, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0728, + "step": 35967 + }, + { + "epoch": 28.728434504792332, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 35968 + }, + { + "epoch": 28.72923322683706, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 35969 + }, + { + "epoch": 28.730031948881788, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 35970 + }, + { + "epoch": 28.73083067092652, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 35971 + }, + { + "epoch": 28.731629392971247, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 35972 + }, + { + "epoch": 28.732428115015974, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 35973 + }, + { + "epoch": 28.733226837060702, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 35974 + }, + { + "epoch": 28.734025559105433, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 35975 + }, + { + "epoch": 28.73482428115016, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 35976 + }, + { + "epoch": 28.73562300319489, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 35977 + }, + { + "epoch": 28.736421725239616, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 35978 + }, + { + "epoch": 28.737220447284344, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 35979 + }, + { + "epoch": 28.738019169329075, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 35980 + }, + { + "epoch": 28.738817891373802, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 35981 + }, + { + "epoch": 28.73961661341853, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 35982 + }, + { + "epoch": 28.740415335463258, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 35983 + }, + { + "epoch": 28.74121405750799, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 35984 + }, + { + "epoch": 28.742012779552716, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 35985 + }, + { + "epoch": 28.742811501597444, + "grad_norm": 0.3125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 35986 + }, + { + "epoch": 28.74361022364217, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 35987 + }, + { + "epoch": 28.7444089456869, + "grad_norm": 0.3046875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 35988 + }, + { + "epoch": 28.74520766773163, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 35989 + }, + { + "epoch": 28.74600638977636, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 35990 + }, + { + "epoch": 28.746805111821086, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 35991 + }, + { + "epoch": 28.747603833865814, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 35992 + }, + { + "epoch": 28.748402555910545, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 35993 + }, + { + "epoch": 28.749201277955272, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 35994 + }, + { + "epoch": 28.75, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 35995 + }, + { + "epoch": 28.750798722044728, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 35996 + }, + { + "epoch": 28.751597444089455, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 35997 + }, + { + "epoch": 28.752396166134186, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 35998 + }, + { + "epoch": 28.753194888178914, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 35999 + }, + { + "epoch": 28.75399361022364, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 36000 + }, + { + "epoch": 28.75479233226837, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 36001 + }, + { + "epoch": 28.7555910543131, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 36002 + }, + { + "epoch": 28.75638977635783, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 36003 + }, + { + "epoch": 28.757188498402556, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 36004 + }, + { + "epoch": 28.757987220447284, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36005 + }, + { + "epoch": 28.75878594249201, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 36006 + }, + { + "epoch": 28.759584664536742, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 36007 + }, + { + "epoch": 28.76038338658147, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 36008 + }, + { + "epoch": 28.761182108626198, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 36009 + }, + { + "epoch": 28.761980830670925, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 36010 + }, + { + "epoch": 28.762779552715656, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 36011 + }, + { + "epoch": 28.763578274760384, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 36012 + }, + { + "epoch": 28.76437699680511, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 36013 + }, + { + "epoch": 28.76517571884984, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 36014 + }, + { + "epoch": 28.765974440894567, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 36015 + }, + { + "epoch": 28.766773162939298, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 36016 + }, + { + "epoch": 28.767571884984026, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 36017 + }, + { + "epoch": 28.768370607028753, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 36018 + }, + { + "epoch": 28.76916932907348, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 36019 + }, + { + "epoch": 28.769968051118212, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 36020 + }, + { + "epoch": 28.77076677316294, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 36021 + }, + { + "epoch": 28.771565495207668, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 36022 + }, + { + "epoch": 28.772364217252395, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 36023 + }, + { + "epoch": 28.773162939297123, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0748, + "step": 36024 + }, + { + "epoch": 28.773961661341854, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 36025 + }, + { + "epoch": 28.77476038338658, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 36026 + }, + { + "epoch": 28.77555910543131, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 36027 + }, + { + "epoch": 28.776357827476037, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 36028 + }, + { + "epoch": 28.777156549520768, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 36029 + }, + { + "epoch": 28.777955271565496, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 36030 + }, + { + "epoch": 28.778753993610223, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36031 + }, + { + "epoch": 28.77955271565495, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 36032 + }, + { + "epoch": 28.78035143769968, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0999, + "step": 36033 + }, + { + "epoch": 28.78115015974441, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 36034 + }, + { + "epoch": 28.781948881789138, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 36035 + }, + { + "epoch": 28.782747603833865, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 36036 + }, + { + "epoch": 28.783546325878593, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 36037 + }, + { + "epoch": 28.784345047923324, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 36038 + }, + { + "epoch": 28.78514376996805, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 36039 + }, + { + "epoch": 28.78594249201278, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 36040 + }, + { + "epoch": 28.786741214057507, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 36041 + }, + { + "epoch": 28.787539936102235, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 36042 + }, + { + "epoch": 28.788338658146966, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 36043 + }, + { + "epoch": 28.789137380191693, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 36044 + }, + { + "epoch": 28.78993610223642, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 36045 + }, + { + "epoch": 28.79073482428115, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 36046 + }, + { + "epoch": 28.79153354632588, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 36047 + }, + { + "epoch": 28.792332268370608, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 36048 + }, + { + "epoch": 28.793130990415335, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 36049 + }, + { + "epoch": 28.793929712460063, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 36050 + }, + { + "epoch": 28.794728434504794, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 36051 + }, + { + "epoch": 28.79552715654952, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 36052 + }, + { + "epoch": 28.79632587859425, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 36053 + }, + { + "epoch": 28.797124600638977, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 36054 + }, + { + "epoch": 28.797923322683705, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 36055 + }, + { + "epoch": 28.798722044728436, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 36056 + }, + { + "epoch": 28.799520766773163, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 36057 + }, + { + "epoch": 28.80031948881789, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 36058 + }, + { + "epoch": 28.80111821086262, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 36059 + }, + { + "epoch": 28.80191693290735, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 36060 + }, + { + "epoch": 28.802715654952078, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 36061 + }, + { + "epoch": 28.803514376996805, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 36062 + }, + { + "epoch": 28.804313099041533, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 36063 + }, + { + "epoch": 28.80511182108626, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 36064 + }, + { + "epoch": 28.80591054313099, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 36065 + }, + { + "epoch": 28.80670926517572, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 36066 + }, + { + "epoch": 28.807507987220447, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 36067 + }, + { + "epoch": 28.808306709265175, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 36068 + }, + { + "epoch": 28.809105431309906, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 36069 + }, + { + "epoch": 28.809904153354633, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 36070 + }, + { + "epoch": 28.81070287539936, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 36071 + }, + { + "epoch": 28.81150159744409, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 36072 + }, + { + "epoch": 28.812300319488816, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 36073 + }, + { + "epoch": 28.813099041533548, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36074 + }, + { + "epoch": 28.813897763578275, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 36075 + }, + { + "epoch": 28.814696485623003, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 36076 + }, + { + "epoch": 28.81549520766773, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 36077 + }, + { + "epoch": 28.81629392971246, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 36078 + }, + { + "epoch": 28.81709265175719, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 36079 + }, + { + "epoch": 28.817891373801917, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 36080 + }, + { + "epoch": 28.818690095846645, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 36081 + }, + { + "epoch": 28.819488817891372, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0992, + "step": 36082 + }, + { + "epoch": 28.820287539936103, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 36083 + }, + { + "epoch": 28.82108626198083, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 36084 + }, + { + "epoch": 28.82188498402556, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.1012, + "step": 36085 + }, + { + "epoch": 28.822683706070286, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36086 + }, + { + "epoch": 28.823482428115017, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 36087 + }, + { + "epoch": 28.824281150159745, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 36088 + }, + { + "epoch": 28.825079872204473, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36089 + }, + { + "epoch": 28.8258785942492, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 36090 + }, + { + "epoch": 28.826677316293928, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 36091 + }, + { + "epoch": 28.82747603833866, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 36092 + }, + { + "epoch": 28.828274760383387, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 36093 + }, + { + "epoch": 28.829073482428115, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 36094 + }, + { + "epoch": 28.829872204472842, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 36095 + }, + { + "epoch": 28.830670926517573, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36096 + }, + { + "epoch": 28.8314696485623, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 36097 + }, + { + "epoch": 28.83226837060703, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 36098 + }, + { + "epoch": 28.833067092651756, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 36099 + }, + { + "epoch": 28.833865814696484, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 36100 + }, + { + "epoch": 28.834664536741215, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 36101 + }, + { + "epoch": 28.835463258785943, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36102 + }, + { + "epoch": 28.83626198083067, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36103 + }, + { + "epoch": 28.837060702875398, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 36104 + }, + { + "epoch": 28.83785942492013, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 36105 + }, + { + "epoch": 28.838658146964857, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 36106 + }, + { + "epoch": 28.839456869009584, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 36107 + }, + { + "epoch": 28.840255591054312, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 36108 + }, + { + "epoch": 28.841054313099043, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 36109 + }, + { + "epoch": 28.84185303514377, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 36110 + }, + { + "epoch": 28.8426517571885, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 36111 + }, + { + "epoch": 28.843450479233226, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 36112 + }, + { + "epoch": 28.844249201277954, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 36113 + }, + { + "epoch": 28.845047923322685, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 36114 + }, + { + "epoch": 28.845846645367413, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 36115 + }, + { + "epoch": 28.84664536741214, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 36116 + }, + { + "epoch": 28.847444089456868, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 36117 + }, + { + "epoch": 28.8482428115016, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 36118 + }, + { + "epoch": 28.849041533546327, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0758, + "step": 36119 + }, + { + "epoch": 28.849840255591054, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 36120 + }, + { + "epoch": 28.850638977635782, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 36121 + }, + { + "epoch": 28.85143769968051, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 36122 + }, + { + "epoch": 28.85223642172524, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 36123 + }, + { + "epoch": 28.85303514376997, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 36124 + }, + { + "epoch": 28.853833865814696, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 36125 + }, + { + "epoch": 28.854632587859424, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 36126 + }, + { + "epoch": 28.855431309904155, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 36127 + }, + { + "epoch": 28.856230031948883, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 36128 + }, + { + "epoch": 28.85702875399361, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 36129 + }, + { + "epoch": 28.857827476038338, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 36130 + }, + { + "epoch": 28.858626198083066, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36131 + }, + { + "epoch": 28.859424920127797, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 36132 + }, + { + "epoch": 28.860223642172524, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 36133 + }, + { + "epoch": 28.861022364217252, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 36134 + }, + { + "epoch": 28.86182108626198, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 36135 + }, + { + "epoch": 28.86261980830671, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 36136 + }, + { + "epoch": 28.86341853035144, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36137 + }, + { + "epoch": 28.864217252396166, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 36138 + }, + { + "epoch": 28.865015974440894, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 36139 + }, + { + "epoch": 28.86581469648562, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 36140 + }, + { + "epoch": 28.866613418530353, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 36141 + }, + { + "epoch": 28.86741214057508, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 36142 + }, + { + "epoch": 28.868210862619808, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 36143 + }, + { + "epoch": 28.869009584664536, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 36144 + }, + { + "epoch": 28.869808306709267, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 36145 + }, + { + "epoch": 28.870607028753994, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36146 + }, + { + "epoch": 28.871405750798722, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 36147 + }, + { + "epoch": 28.87220447284345, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 36148 + }, + { + "epoch": 28.873003194888177, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 36149 + }, + { + "epoch": 28.87380191693291, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 36150 + }, + { + "epoch": 28.874600638977636, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 36151 + }, + { + "epoch": 28.875399361022364, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 36152 + }, + { + "epoch": 28.87619808306709, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 36153 + }, + { + "epoch": 28.876996805111823, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 36154 + }, + { + "epoch": 28.87779552715655, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 36155 + }, + { + "epoch": 28.878594249201278, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 36156 + }, + { + "epoch": 28.879392971246006, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36157 + }, + { + "epoch": 28.880191693290733, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 36158 + }, + { + "epoch": 28.880990415335464, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 36159 + }, + { + "epoch": 28.881789137380192, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 36160 + }, + { + "epoch": 28.88258785942492, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 36161 + }, + { + "epoch": 28.883386581469647, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 36162 + }, + { + "epoch": 28.88418530351438, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 36163 + }, + { + "epoch": 28.884984025559106, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 36164 + }, + { + "epoch": 28.885782747603834, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 36165 + }, + { + "epoch": 28.88658146964856, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 36166 + }, + { + "epoch": 28.88738019169329, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36167 + }, + { + "epoch": 28.88817891373802, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 36168 + }, + { + "epoch": 28.888977635782748, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 36169 + }, + { + "epoch": 28.889776357827476, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 36170 + }, + { + "epoch": 28.890575079872203, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 36171 + }, + { + "epoch": 28.891373801916934, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 36172 + }, + { + "epoch": 28.892172523961662, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 36173 + }, + { + "epoch": 28.89297124600639, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 36174 + }, + { + "epoch": 28.893769968051117, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 36175 + }, + { + "epoch": 28.894568690095845, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 36176 + }, + { + "epoch": 28.895367412140576, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 36177 + }, + { + "epoch": 28.896166134185304, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 36178 + }, + { + "epoch": 28.89696485623003, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36179 + }, + { + "epoch": 28.89776357827476, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 36180 + }, + { + "epoch": 28.89856230031949, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 36181 + }, + { + "epoch": 28.899361022364218, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 36182 + }, + { + "epoch": 28.900159744408946, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 36183 + }, + { + "epoch": 28.900958466453673, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 36184 + }, + { + "epoch": 28.9017571884984, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 36185 + }, + { + "epoch": 28.902555910543132, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 36186 + }, + { + "epoch": 28.90335463258786, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36187 + }, + { + "epoch": 28.904153354632587, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 36188 + }, + { + "epoch": 28.904952076677315, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 36189 + }, + { + "epoch": 28.905750798722046, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 36190 + }, + { + "epoch": 28.906549520766774, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 36191 + }, + { + "epoch": 28.9073482428115, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 36192 + }, + { + "epoch": 28.90814696485623, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 36193 + }, + { + "epoch": 28.908945686900957, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 36194 + }, + { + "epoch": 28.909744408945688, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 36195 + }, + { + "epoch": 28.910543130990416, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 36196 + }, + { + "epoch": 28.911341853035143, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 36197 + }, + { + "epoch": 28.91214057507987, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 36198 + }, + { + "epoch": 28.912939297124602, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 36199 + }, + { + "epoch": 28.91373801916933, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 36200 + }, + { + "epoch": 28.914536741214057, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 36201 + }, + { + "epoch": 28.915335463258785, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 36202 + }, + { + "epoch": 28.916134185303516, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0757, + "step": 36203 + }, + { + "epoch": 28.916932907348244, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 36204 + }, + { + "epoch": 28.91773162939297, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 36205 + }, + { + "epoch": 28.9185303514377, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 36206 + }, + { + "epoch": 28.919329073482427, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 36207 + }, + { + "epoch": 28.920127795527158, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 36208 + }, + { + "epoch": 28.920926517571885, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 36209 + }, + { + "epoch": 28.921725239616613, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 36210 + }, + { + "epoch": 28.92252396166134, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 36211 + }, + { + "epoch": 28.923322683706072, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 36212 + }, + { + "epoch": 28.9241214057508, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 36213 + }, + { + "epoch": 28.924920127795527, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 36214 + }, + { + "epoch": 28.925718849840255, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 36215 + }, + { + "epoch": 28.926517571884983, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 36216 + }, + { + "epoch": 28.927316293929714, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 36217 + }, + { + "epoch": 28.92811501597444, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 36218 + }, + { + "epoch": 28.92891373801917, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 36219 + }, + { + "epoch": 28.929712460063897, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 36220 + }, + { + "epoch": 28.930511182108628, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 36221 + }, + { + "epoch": 28.931309904153355, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 36222 + }, + { + "epoch": 28.932108626198083, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 36223 + }, + { + "epoch": 28.93290734824281, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 36224 + }, + { + "epoch": 28.93370607028754, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 36225 + }, + { + "epoch": 28.93450479233227, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36226 + }, + { + "epoch": 28.935303514376997, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 36227 + }, + { + "epoch": 28.936102236421725, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 36228 + }, + { + "epoch": 28.936900958466452, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 36229 + }, + { + "epoch": 28.937699680511184, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 36230 + }, + { + "epoch": 28.93849840255591, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 36231 + }, + { + "epoch": 28.93929712460064, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 36232 + }, + { + "epoch": 28.940095846645367, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 36233 + }, + { + "epoch": 28.940894568690094, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 36234 + }, + { + "epoch": 28.941693290734825, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 36235 + }, + { + "epoch": 28.942492012779553, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 36236 + }, + { + "epoch": 28.94329073482428, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 36237 + }, + { + "epoch": 28.94408945686901, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 36238 + }, + { + "epoch": 28.94488817891374, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 36239 + }, + { + "epoch": 28.945686900958467, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36240 + }, + { + "epoch": 28.946485623003195, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 36241 + }, + { + "epoch": 28.947284345047922, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 36242 + }, + { + "epoch": 28.94808306709265, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 36243 + }, + { + "epoch": 28.94888178913738, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 36244 + }, + { + "epoch": 28.94968051118211, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 36245 + }, + { + "epoch": 28.950479233226837, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 36246 + }, + { + "epoch": 28.951277955271564, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 36247 + }, + { + "epoch": 28.952076677316295, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36248 + }, + { + "epoch": 28.952875399361023, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36249 + }, + { + "epoch": 28.95367412140575, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 36250 + }, + { + "epoch": 28.95447284345048, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 36251 + }, + { + "epoch": 28.955271565495206, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 36252 + }, + { + "epoch": 28.956070287539937, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 36253 + }, + { + "epoch": 28.956869009584665, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 36254 + }, + { + "epoch": 28.957667731629392, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 36255 + }, + { + "epoch": 28.95846645367412, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0754, + "step": 36256 + }, + { + "epoch": 28.95926517571885, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 36257 + }, + { + "epoch": 28.96006389776358, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 36258 + }, + { + "epoch": 28.960862619808307, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 36259 + }, + { + "epoch": 28.961661341853034, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 36260 + }, + { + "epoch": 28.962460063897765, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 36261 + }, + { + "epoch": 28.963258785942493, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 36262 + }, + { + "epoch": 28.96405750798722, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36263 + }, + { + "epoch": 28.96485623003195, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 36264 + }, + { + "epoch": 28.965654952076676, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 36265 + }, + { + "epoch": 28.966453674121407, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36266 + }, + { + "epoch": 28.967252396166135, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36267 + }, + { + "epoch": 28.968051118210862, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 36268 + }, + { + "epoch": 28.96884984025559, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36269 + }, + { + "epoch": 28.96964856230032, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 36270 + }, + { + "epoch": 28.97044728434505, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 36271 + }, + { + "epoch": 28.971246006389777, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 36272 + }, + { + "epoch": 28.972044728434504, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 36273 + }, + { + "epoch": 28.972843450479232, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 36274 + }, + { + "epoch": 28.973642172523963, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 36275 + }, + { + "epoch": 28.97444089456869, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 36276 + }, + { + "epoch": 28.97523961661342, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 36277 + }, + { + "epoch": 28.976038338658146, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 36278 + }, + { + "epoch": 28.976837060702877, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 36279 + }, + { + "epoch": 28.977635782747605, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 36280 + }, + { + "epoch": 28.978434504792332, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 36281 + }, + { + "epoch": 28.97923322683706, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 36282 + }, + { + "epoch": 28.980031948881788, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 36283 + }, + { + "epoch": 28.98083067092652, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 36284 + }, + { + "epoch": 28.981629392971247, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 36285 + }, + { + "epoch": 28.982428115015974, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 36286 + }, + { + "epoch": 28.983226837060702, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 36287 + }, + { + "epoch": 28.984025559105433, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 36288 + }, + { + "epoch": 28.98482428115016, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 36289 + }, + { + "epoch": 28.98562300319489, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 36290 + }, + { + "epoch": 28.986421725239616, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 36291 + }, + { + "epoch": 28.987220447284344, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36292 + }, + { + "epoch": 28.988019169329075, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 36293 + }, + { + "epoch": 28.988817891373802, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36294 + }, + { + "epoch": 28.98961661341853, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 36295 + }, + { + "epoch": 28.990415335463258, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 36296 + }, + { + "epoch": 28.99121405750799, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36297 + }, + { + "epoch": 28.992012779552716, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 36298 + }, + { + "epoch": 28.992811501597444, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36299 + }, + { + "epoch": 28.99361022364217, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 36300 + }, + { + "epoch": 28.9944089456869, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 36301 + }, + { + "epoch": 28.99520766773163, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 36302 + }, + { + "epoch": 28.99600638977636, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 36303 + }, + { + "epoch": 28.996805111821086, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36304 + }, + { + "epoch": 28.997603833865814, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 36305 + }, + { + "epoch": 28.998402555910545, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 36306 + }, + { + "epoch": 28.999201277955272, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 36307 + }, + { + "epoch": 29.0, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 36308 + }, + { + "epoch": 29.000798722044728, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36309 + }, + { + "epoch": 29.001597444089455, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 36310 + }, + { + "epoch": 29.002396166134186, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 36311 + }, + { + "epoch": 29.003194888178914, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 36312 + }, + { + "epoch": 29.00399361022364, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 36313 + }, + { + "epoch": 29.00479233226837, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36314 + }, + { + "epoch": 29.0055910543131, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 36315 + }, + { + "epoch": 29.00638977635783, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 36316 + }, + { + "epoch": 29.007188498402556, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 36317 + }, + { + "epoch": 29.007987220447284, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 36318 + }, + { + "epoch": 29.00878594249201, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36319 + }, + { + "epoch": 29.009584664536742, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 36320 + }, + { + "epoch": 29.01038338658147, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 36321 + }, + { + "epoch": 29.011182108626198, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 36322 + }, + { + "epoch": 29.011980830670925, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 36323 + }, + { + "epoch": 29.012779552715656, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 36324 + }, + { + "epoch": 29.013578274760384, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0745, + "step": 36325 + }, + { + "epoch": 29.01437699680511, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 36326 + }, + { + "epoch": 29.01517571884984, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 36327 + }, + { + "epoch": 29.015974440894567, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 36328 + }, + { + "epoch": 29.016773162939298, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 36329 + }, + { + "epoch": 29.017571884984026, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36330 + }, + { + "epoch": 29.018370607028753, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36331 + }, + { + "epoch": 29.01916932907348, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 36332 + }, + { + "epoch": 29.019968051118212, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 36333 + }, + { + "epoch": 29.02076677316294, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36334 + }, + { + "epoch": 29.021565495207668, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 36335 + }, + { + "epoch": 29.022364217252395, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36336 + }, + { + "epoch": 29.023162939297123, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 36337 + }, + { + "epoch": 29.023961661341854, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 36338 + }, + { + "epoch": 29.02476038338658, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 36339 + }, + { + "epoch": 29.02555910543131, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0743, + "step": 36340 + }, + { + "epoch": 29.026357827476037, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 36341 + }, + { + "epoch": 29.027156549520768, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 36342 + }, + { + "epoch": 29.027955271565496, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 36343 + }, + { + "epoch": 29.028753993610223, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 36344 + }, + { + "epoch": 29.02955271565495, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 36345 + }, + { + "epoch": 29.03035143769968, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 36346 + }, + { + "epoch": 29.03115015974441, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 36347 + }, + { + "epoch": 29.031948881789138, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 36348 + }, + { + "epoch": 29.032747603833865, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 36349 + }, + { + "epoch": 29.033546325878593, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 36350 + }, + { + "epoch": 29.034345047923324, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 36351 + }, + { + "epoch": 29.03514376996805, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 36352 + }, + { + "epoch": 29.03594249201278, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 36353 + }, + { + "epoch": 29.036741214057507, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 36354 + }, + { + "epoch": 29.037539936102238, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 36355 + }, + { + "epoch": 29.038338658146966, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 36356 + }, + { + "epoch": 29.039137380191693, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 36357 + }, + { + "epoch": 29.03993610223642, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 36358 + }, + { + "epoch": 29.04073482428115, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 36359 + }, + { + "epoch": 29.04153354632588, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 36360 + }, + { + "epoch": 29.042332268370608, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 36361 + }, + { + "epoch": 29.043130990415335, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 36362 + }, + { + "epoch": 29.043929712460063, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 36363 + }, + { + "epoch": 29.044728434504794, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 36364 + }, + { + "epoch": 29.04552715654952, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 36365 + }, + { + "epoch": 29.04632587859425, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 36366 + }, + { + "epoch": 29.047124600638977, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 36367 + }, + { + "epoch": 29.047923322683705, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 36368 + }, + { + "epoch": 29.048722044728436, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 36369 + }, + { + "epoch": 29.049520766773163, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 36370 + }, + { + "epoch": 29.05031948881789, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 36371 + }, + { + "epoch": 29.05111821086262, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 36372 + }, + { + "epoch": 29.05191693290735, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 36373 + }, + { + "epoch": 29.052715654952078, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 36374 + }, + { + "epoch": 29.053514376996805, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36375 + }, + { + "epoch": 29.054313099041533, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 36376 + }, + { + "epoch": 29.05511182108626, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 36377 + }, + { + "epoch": 29.05591054313099, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 36378 + }, + { + "epoch": 29.05670926517572, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 36379 + }, + { + "epoch": 29.057507987220447, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 36380 + }, + { + "epoch": 29.058306709265175, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 36381 + }, + { + "epoch": 29.059105431309906, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 36382 + }, + { + "epoch": 29.059904153354633, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 36383 + }, + { + "epoch": 29.06070287539936, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 36384 + }, + { + "epoch": 29.06150159744409, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 36385 + }, + { + "epoch": 29.062300319488816, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 36386 + }, + { + "epoch": 29.063099041533548, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 36387 + }, + { + "epoch": 29.063897763578275, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 36388 + }, + { + "epoch": 29.064696485623003, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 36389 + }, + { + "epoch": 29.06549520766773, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36390 + }, + { + "epoch": 29.06629392971246, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 36391 + }, + { + "epoch": 29.06709265175719, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36392 + }, + { + "epoch": 29.067891373801917, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 36393 + }, + { + "epoch": 29.068690095846645, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 36394 + }, + { + "epoch": 29.069488817891372, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 36395 + }, + { + "epoch": 29.070287539936103, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 36396 + }, + { + "epoch": 29.07108626198083, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 36397 + }, + { + "epoch": 29.07188498402556, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 36398 + }, + { + "epoch": 29.072683706070286, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36399 + }, + { + "epoch": 29.073482428115017, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36400 + }, + { + "epoch": 29.074281150159745, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 36401 + }, + { + "epoch": 29.075079872204473, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 36402 + }, + { + "epoch": 29.0758785942492, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 36403 + }, + { + "epoch": 29.076677316293928, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 36404 + }, + { + "epoch": 29.07747603833866, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 36405 + }, + { + "epoch": 29.078274760383387, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 36406 + }, + { + "epoch": 29.079073482428115, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 36407 + }, + { + "epoch": 29.079872204472842, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 36408 + }, + { + "epoch": 29.080670926517573, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36409 + }, + { + "epoch": 29.0814696485623, + "grad_norm": 0.298828125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 36410 + }, + { + "epoch": 29.08226837060703, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 36411 + }, + { + "epoch": 29.083067092651756, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 36412 + }, + { + "epoch": 29.083865814696484, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 36413 + }, + { + "epoch": 29.084664536741215, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 36414 + }, + { + "epoch": 29.085463258785943, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36415 + }, + { + "epoch": 29.08626198083067, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 36416 + }, + { + "epoch": 29.087060702875398, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 36417 + }, + { + "epoch": 29.08785942492013, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 36418 + }, + { + "epoch": 29.088658146964857, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 36419 + }, + { + "epoch": 29.089456869009584, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 36420 + }, + { + "epoch": 29.090255591054312, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 36421 + }, + { + "epoch": 29.09105431309904, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 36422 + }, + { + "epoch": 29.09185303514377, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 36423 + }, + { + "epoch": 29.0926517571885, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 36424 + }, + { + "epoch": 29.093450479233226, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 36425 + }, + { + "epoch": 29.094249201277954, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 36426 + }, + { + "epoch": 29.095047923322685, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 36427 + }, + { + "epoch": 29.095846645367413, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 36428 + }, + { + "epoch": 29.09664536741214, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 36429 + }, + { + "epoch": 29.097444089456868, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 36430 + }, + { + "epoch": 29.0982428115016, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36431 + }, + { + "epoch": 29.099041533546327, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 36432 + }, + { + "epoch": 29.099840255591054, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 36433 + }, + { + "epoch": 29.100638977635782, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 36434 + }, + { + "epoch": 29.10143769968051, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 36435 + }, + { + "epoch": 29.10223642172524, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 36436 + }, + { + "epoch": 29.10303514376997, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 36437 + }, + { + "epoch": 29.103833865814696, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 36438 + }, + { + "epoch": 29.104632587859424, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36439 + }, + { + "epoch": 29.105431309904155, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 36440 + }, + { + "epoch": 29.106230031948883, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 36441 + }, + { + "epoch": 29.10702875399361, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 36442 + }, + { + "epoch": 29.107827476038338, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 36443 + }, + { + "epoch": 29.108626198083066, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 36444 + }, + { + "epoch": 29.109424920127797, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 36445 + }, + { + "epoch": 29.110223642172524, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 36446 + }, + { + "epoch": 29.111022364217252, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 36447 + }, + { + "epoch": 29.11182108626198, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36448 + }, + { + "epoch": 29.11261980830671, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 36449 + }, + { + "epoch": 29.11341853035144, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 36450 + }, + { + "epoch": 29.114217252396166, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 36451 + }, + { + "epoch": 29.115015974440894, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 36452 + }, + { + "epoch": 29.11581469648562, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 36453 + }, + { + "epoch": 29.116613418530353, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 36454 + }, + { + "epoch": 29.11741214057508, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 36455 + }, + { + "epoch": 29.118210862619808, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 36456 + }, + { + "epoch": 29.119009584664536, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 36457 + }, + { + "epoch": 29.119808306709267, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 36458 + }, + { + "epoch": 29.120607028753994, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 36459 + }, + { + "epoch": 29.121405750798722, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 36460 + }, + { + "epoch": 29.12220447284345, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 36461 + }, + { + "epoch": 29.123003194888177, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 36462 + }, + { + "epoch": 29.12380191693291, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 36463 + }, + { + "epoch": 29.124600638977636, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 36464 + }, + { + "epoch": 29.125399361022364, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 36465 + }, + { + "epoch": 29.12619808306709, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36466 + }, + { + "epoch": 29.126996805111823, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 36467 + }, + { + "epoch": 29.12779552715655, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1007, + "step": 36468 + }, + { + "epoch": 29.128594249201278, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 36469 + }, + { + "epoch": 29.129392971246006, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 36470 + }, + { + "epoch": 29.130191693290733, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36471 + }, + { + "epoch": 29.130990415335464, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 36472 + }, + { + "epoch": 29.131789137380192, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 36473 + }, + { + "epoch": 29.13258785942492, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 36474 + }, + { + "epoch": 29.133386581469647, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.1, + "step": 36475 + }, + { + "epoch": 29.13418530351438, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 36476 + }, + { + "epoch": 29.134984025559106, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 36477 + }, + { + "epoch": 29.135782747603834, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 36478 + }, + { + "epoch": 29.13658146964856, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 36479 + }, + { + "epoch": 29.13738019169329, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 36480 + }, + { + "epoch": 29.13817891373802, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 36481 + }, + { + "epoch": 29.138977635782748, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 36482 + }, + { + "epoch": 29.139776357827476, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 36483 + }, + { + "epoch": 29.140575079872203, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 36484 + }, + { + "epoch": 29.141373801916934, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 36485 + }, + { + "epoch": 29.142172523961662, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 36486 + }, + { + "epoch": 29.14297124600639, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 36487 + }, + { + "epoch": 29.143769968051117, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 36488 + }, + { + "epoch": 29.144568690095845, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.1019, + "step": 36489 + }, + { + "epoch": 29.145367412140576, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36490 + }, + { + "epoch": 29.146166134185304, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 36491 + }, + { + "epoch": 29.14696485623003, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36492 + }, + { + "epoch": 29.14776357827476, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 36493 + }, + { + "epoch": 29.14856230031949, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 36494 + }, + { + "epoch": 29.149361022364218, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 36495 + }, + { + "epoch": 29.150159744408946, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 36496 + }, + { + "epoch": 29.150958466453673, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 36497 + }, + { + "epoch": 29.1517571884984, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 36498 + }, + { + "epoch": 29.152555910543132, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 36499 + }, + { + "epoch": 29.15335463258786, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 36500 + }, + { + "epoch": 29.154153354632587, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 36501 + }, + { + "epoch": 29.154952076677315, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 36502 + }, + { + "epoch": 29.155750798722046, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 36503 + }, + { + "epoch": 29.156549520766774, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36504 + }, + { + "epoch": 29.1573482428115, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 36505 + }, + { + "epoch": 29.15814696485623, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36506 + }, + { + "epoch": 29.15894568690096, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 36507 + }, + { + "epoch": 29.159744408945688, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 36508 + }, + { + "epoch": 29.160543130990416, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 36509 + }, + { + "epoch": 29.161341853035143, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 36510 + }, + { + "epoch": 29.16214057507987, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36511 + }, + { + "epoch": 29.162939297124602, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 36512 + }, + { + "epoch": 29.16373801916933, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 36513 + }, + { + "epoch": 29.164536741214057, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36514 + }, + { + "epoch": 29.165335463258785, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 36515 + }, + { + "epoch": 29.166134185303516, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 36516 + }, + { + "epoch": 29.166932907348244, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 36517 + }, + { + "epoch": 29.16773162939297, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 36518 + }, + { + "epoch": 29.1685303514377, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 36519 + }, + { + "epoch": 29.169329073482427, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 36520 + }, + { + "epoch": 29.170127795527158, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 36521 + }, + { + "epoch": 29.170926517571885, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 36522 + }, + { + "epoch": 29.171725239616613, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 36523 + }, + { + "epoch": 29.17252396166134, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 36524 + }, + { + "epoch": 29.173322683706072, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0998, + "step": 36525 + }, + { + "epoch": 29.1741214057508, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 36526 + }, + { + "epoch": 29.174920127795527, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 36527 + }, + { + "epoch": 29.175718849840255, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 36528 + }, + { + "epoch": 29.176517571884983, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 36529 + }, + { + "epoch": 29.177316293929714, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 36530 + }, + { + "epoch": 29.17811501597444, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 36531 + }, + { + "epoch": 29.17891373801917, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 36532 + }, + { + "epoch": 29.179712460063897, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 36533 + }, + { + "epoch": 29.180511182108628, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 36534 + }, + { + "epoch": 29.181309904153355, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 36535 + }, + { + "epoch": 29.182108626198083, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 36536 + }, + { + "epoch": 29.18290734824281, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 36537 + }, + { + "epoch": 29.18370607028754, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 36538 + }, + { + "epoch": 29.18450479233227, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 36539 + }, + { + "epoch": 29.185303514376997, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 36540 + }, + { + "epoch": 29.186102236421725, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 36541 + }, + { + "epoch": 29.186900958466452, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 36542 + }, + { + "epoch": 29.187699680511184, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36543 + }, + { + "epoch": 29.18849840255591, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 36544 + }, + { + "epoch": 29.18929712460064, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 36545 + }, + { + "epoch": 29.190095846645367, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 36546 + }, + { + "epoch": 29.190894568690094, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 36547 + }, + { + "epoch": 29.191693290734825, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.075, + "step": 36548 + }, + { + "epoch": 29.192492012779553, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 36549 + }, + { + "epoch": 29.19329073482428, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0733, + "step": 36550 + }, + { + "epoch": 29.19408945686901, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 36551 + }, + { + "epoch": 29.19488817891374, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 36552 + }, + { + "epoch": 29.195686900958467, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 36553 + }, + { + "epoch": 29.196485623003195, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 36554 + }, + { + "epoch": 29.197284345047922, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 36555 + }, + { + "epoch": 29.19808306709265, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 36556 + }, + { + "epoch": 29.19888178913738, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 36557 + }, + { + "epoch": 29.19968051118211, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0727, + "step": 36558 + }, + { + "epoch": 29.200479233226837, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 36559 + }, + { + "epoch": 29.201277955271564, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 36560 + }, + { + "epoch": 29.202076677316295, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 36561 + }, + { + "epoch": 29.202875399361023, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 36562 + }, + { + "epoch": 29.20367412140575, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 36563 + }, + { + "epoch": 29.20447284345048, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 36564 + }, + { + "epoch": 29.205271565495206, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 36565 + }, + { + "epoch": 29.206070287539937, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 36566 + }, + { + "epoch": 29.206869009584665, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 36567 + }, + { + "epoch": 29.207667731629392, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 36568 + }, + { + "epoch": 29.20846645367412, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 36569 + }, + { + "epoch": 29.20926517571885, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 36570 + }, + { + "epoch": 29.21006389776358, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 36571 + }, + { + "epoch": 29.210862619808307, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 36572 + }, + { + "epoch": 29.211661341853034, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 36573 + }, + { + "epoch": 29.212460063897762, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36574 + }, + { + "epoch": 29.213258785942493, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 36575 + }, + { + "epoch": 29.21405750798722, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 36576 + }, + { + "epoch": 29.21485623003195, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 36577 + }, + { + "epoch": 29.215654952076676, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 36578 + }, + { + "epoch": 29.216453674121407, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36579 + }, + { + "epoch": 29.217252396166135, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.076, + "step": 36580 + }, + { + "epoch": 29.218051118210862, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 36581 + }, + { + "epoch": 29.21884984025559, + "grad_norm": 0.10302734375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 36582 + }, + { + "epoch": 29.21964856230032, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 36583 + }, + { + "epoch": 29.22044728434505, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 36584 + }, + { + "epoch": 29.221246006389777, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 36585 + }, + { + "epoch": 29.222044728434504, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 36586 + }, + { + "epoch": 29.222843450479232, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36587 + }, + { + "epoch": 29.223642172523963, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36588 + }, + { + "epoch": 29.22444089456869, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 36589 + }, + { + "epoch": 29.22523961661342, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 36590 + }, + { + "epoch": 29.226038338658146, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 36591 + }, + { + "epoch": 29.226837060702877, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 36592 + }, + { + "epoch": 29.227635782747605, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 36593 + }, + { + "epoch": 29.228434504792332, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 36594 + }, + { + "epoch": 29.22923322683706, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 36595 + }, + { + "epoch": 29.230031948881788, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 36596 + }, + { + "epoch": 29.23083067092652, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 36597 + }, + { + "epoch": 29.231629392971247, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 36598 + }, + { + "epoch": 29.232428115015974, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 36599 + }, + { + "epoch": 29.233226837060702, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 36600 + }, + { + "epoch": 29.234025559105433, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 36601 + }, + { + "epoch": 29.23482428115016, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 36602 + }, + { + "epoch": 29.23562300319489, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 36603 + }, + { + "epoch": 29.236421725239616, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 36604 + }, + { + "epoch": 29.237220447284344, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36605 + }, + { + "epoch": 29.238019169329075, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 36606 + }, + { + "epoch": 29.238817891373802, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 36607 + }, + { + "epoch": 29.23961661341853, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 36608 + }, + { + "epoch": 29.240415335463258, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 36609 + }, + { + "epoch": 29.24121405750799, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 36610 + }, + { + "epoch": 29.242012779552716, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 36611 + }, + { + "epoch": 29.242811501597444, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 36612 + }, + { + "epoch": 29.24361022364217, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 36613 + }, + { + "epoch": 29.2444089456869, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 36614 + }, + { + "epoch": 29.24520766773163, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 36615 + }, + { + "epoch": 29.24600638977636, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 36616 + }, + { + "epoch": 29.246805111821086, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 36617 + }, + { + "epoch": 29.247603833865814, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 36618 + }, + { + "epoch": 29.248402555910545, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 36619 + }, + { + "epoch": 29.249201277955272, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 36620 + }, + { + "epoch": 29.25, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 36621 + }, + { + "epoch": 29.250798722044728, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 36622 + }, + { + "epoch": 29.251597444089455, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 36623 + }, + { + "epoch": 29.252396166134186, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 36624 + }, + { + "epoch": 29.253194888178914, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 36625 + }, + { + "epoch": 29.25399361022364, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 36626 + }, + { + "epoch": 29.25479233226837, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 36627 + }, + { + "epoch": 29.2555910543131, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 36628 + }, + { + "epoch": 29.25638977635783, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 36629 + }, + { + "epoch": 29.257188498402556, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 36630 + }, + { + "epoch": 29.257987220447284, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 36631 + }, + { + "epoch": 29.25878594249201, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 36632 + }, + { + "epoch": 29.259584664536742, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 36633 + }, + { + "epoch": 29.26038338658147, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 36634 + }, + { + "epoch": 29.261182108626198, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 36635 + }, + { + "epoch": 29.261980830670925, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 36636 + }, + { + "epoch": 29.262779552715656, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 36637 + }, + { + "epoch": 29.263578274760384, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 36638 + }, + { + "epoch": 29.26437699680511, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 36639 + }, + { + "epoch": 29.26517571884984, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 36640 + }, + { + "epoch": 29.265974440894567, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0763, + "step": 36641 + }, + { + "epoch": 29.266773162939298, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 36642 + }, + { + "epoch": 29.267571884984026, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36643 + }, + { + "epoch": 29.268370607028753, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 36644 + }, + { + "epoch": 29.26916932907348, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 36645 + }, + { + "epoch": 29.269968051118212, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 36646 + }, + { + "epoch": 29.27076677316294, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 36647 + }, + { + "epoch": 29.271565495207668, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 36648 + }, + { + "epoch": 29.272364217252395, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 36649 + }, + { + "epoch": 29.273162939297123, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 36650 + }, + { + "epoch": 29.273961661341854, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 36651 + }, + { + "epoch": 29.27476038338658, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 36652 + }, + { + "epoch": 29.27555910543131, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 36653 + }, + { + "epoch": 29.276357827476037, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 36654 + }, + { + "epoch": 29.277156549520768, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 36655 + }, + { + "epoch": 29.277955271565496, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 36656 + }, + { + "epoch": 29.278753993610223, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 36657 + }, + { + "epoch": 29.27955271565495, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 36658 + }, + { + "epoch": 29.28035143769968, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 36659 + }, + { + "epoch": 29.28115015974441, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 36660 + }, + { + "epoch": 29.281948881789138, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 36661 + }, + { + "epoch": 29.282747603833865, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 36662 + }, + { + "epoch": 29.283546325878593, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 36663 + }, + { + "epoch": 29.284345047923324, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 36664 + }, + { + "epoch": 29.28514376996805, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 36665 + }, + { + "epoch": 29.28594249201278, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36666 + }, + { + "epoch": 29.286741214057507, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 36667 + }, + { + "epoch": 29.287539936102238, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 36668 + }, + { + "epoch": 29.288338658146966, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 36669 + }, + { + "epoch": 29.289137380191693, + "grad_norm": 0.435546875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 36670 + }, + { + "epoch": 29.28993610223642, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 36671 + }, + { + "epoch": 29.29073482428115, + "grad_norm": 0.7109375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 36672 + }, + { + "epoch": 29.29153354632588, + "grad_norm": 0.2255859375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 36673 + }, + { + "epoch": 29.292332268370608, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 36674 + }, + { + "epoch": 29.293130990415335, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 36675 + }, + { + "epoch": 29.293929712460063, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 36676 + }, + { + "epoch": 29.294728434504794, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 36677 + }, + { + "epoch": 29.29552715654952, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 36678 + }, + { + "epoch": 29.29632587859425, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 36679 + }, + { + "epoch": 29.297124600638977, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 36680 + }, + { + "epoch": 29.297923322683705, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 36681 + }, + { + "epoch": 29.298722044728436, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 36682 + }, + { + "epoch": 29.299520766773163, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 36683 + }, + { + "epoch": 29.30031948881789, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 36684 + }, + { + "epoch": 29.30111821086262, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 36685 + }, + { + "epoch": 29.30191693290735, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.076, + "step": 36686 + }, + { + "epoch": 29.302715654952078, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 36687 + }, + { + "epoch": 29.303514376996805, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 36688 + }, + { + "epoch": 29.304313099041533, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 36689 + }, + { + "epoch": 29.30511182108626, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 36690 + }, + { + "epoch": 29.30591054313099, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 36691 + }, + { + "epoch": 29.30670926517572, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 36692 + }, + { + "epoch": 29.307507987220447, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 36693 + }, + { + "epoch": 29.308306709265175, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36694 + }, + { + "epoch": 29.309105431309906, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 36695 + }, + { + "epoch": 29.309904153354633, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 36696 + }, + { + "epoch": 29.31070287539936, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 36697 + }, + { + "epoch": 29.31150159744409, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 36698 + }, + { + "epoch": 29.312300319488816, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 36699 + }, + { + "epoch": 29.313099041533548, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 36700 + }, + { + "epoch": 29.313897763578275, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 36701 + }, + { + "epoch": 29.314696485623003, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 36702 + }, + { + "epoch": 29.31549520766773, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 36703 + }, + { + "epoch": 29.31629392971246, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 36704 + }, + { + "epoch": 29.31709265175719, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 36705 + }, + { + "epoch": 29.317891373801917, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 36706 + }, + { + "epoch": 29.318690095846645, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 36707 + }, + { + "epoch": 29.319488817891372, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 36708 + }, + { + "epoch": 29.320287539936103, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 36709 + }, + { + "epoch": 29.32108626198083, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 36710 + }, + { + "epoch": 29.32188498402556, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 36711 + }, + { + "epoch": 29.322683706070286, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 36712 + }, + { + "epoch": 29.323482428115017, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 36713 + }, + { + "epoch": 29.324281150159745, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36714 + }, + { + "epoch": 29.325079872204473, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 36715 + }, + { + "epoch": 29.3258785942492, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 36716 + }, + { + "epoch": 29.326677316293928, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 36717 + }, + { + "epoch": 29.32747603833866, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 36718 + }, + { + "epoch": 29.328274760383387, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36719 + }, + { + "epoch": 29.329073482428115, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 36720 + }, + { + "epoch": 29.329872204472842, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 36721 + }, + { + "epoch": 29.330670926517573, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 36722 + }, + { + "epoch": 29.3314696485623, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.1025, + "step": 36723 + }, + { + "epoch": 29.33226837060703, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36724 + }, + { + "epoch": 29.333067092651756, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 36725 + }, + { + "epoch": 29.333865814696484, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 36726 + }, + { + "epoch": 29.334664536741215, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36727 + }, + { + "epoch": 29.335463258785943, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 36728 + }, + { + "epoch": 29.33626198083067, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 36729 + }, + { + "epoch": 29.337060702875398, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 36730 + }, + { + "epoch": 29.33785942492013, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 36731 + }, + { + "epoch": 29.338658146964857, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 36732 + }, + { + "epoch": 29.339456869009584, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 36733 + }, + { + "epoch": 29.340255591054312, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 36734 + }, + { + "epoch": 29.341054313099043, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 36735 + }, + { + "epoch": 29.34185303514377, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 36736 + }, + { + "epoch": 29.3426517571885, + "grad_norm": 0.259765625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 36737 + }, + { + "epoch": 29.343450479233226, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 36738 + }, + { + "epoch": 29.344249201277954, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 36739 + }, + { + "epoch": 29.345047923322685, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 36740 + }, + { + "epoch": 29.345846645367413, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 36741 + }, + { + "epoch": 29.34664536741214, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 36742 + }, + { + "epoch": 29.347444089456868, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 36743 + }, + { + "epoch": 29.3482428115016, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 36744 + }, + { + "epoch": 29.349041533546327, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 36745 + }, + { + "epoch": 29.349840255591054, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 36746 + }, + { + "epoch": 29.350638977635782, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 36747 + }, + { + "epoch": 29.35143769968051, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 36748 + }, + { + "epoch": 29.35223642172524, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 36749 + }, + { + "epoch": 29.35303514376997, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36750 + }, + { + "epoch": 29.353833865814696, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 36751 + }, + { + "epoch": 29.354632587859424, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 36752 + }, + { + "epoch": 29.355431309904155, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 36753 + }, + { + "epoch": 29.356230031948883, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 36754 + }, + { + "epoch": 29.35702875399361, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 36755 + }, + { + "epoch": 29.357827476038338, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0765, + "step": 36756 + }, + { + "epoch": 29.358626198083066, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 36757 + }, + { + "epoch": 29.359424920127797, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 36758 + }, + { + "epoch": 29.360223642172524, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 36759 + }, + { + "epoch": 29.361022364217252, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 36760 + }, + { + "epoch": 29.36182108626198, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 36761 + }, + { + "epoch": 29.36261980830671, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36762 + }, + { + "epoch": 29.36341853035144, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 36763 + }, + { + "epoch": 29.364217252396166, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 36764 + }, + { + "epoch": 29.365015974440894, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 36765 + }, + { + "epoch": 29.36581469648562, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 36766 + }, + { + "epoch": 29.366613418530353, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 36767 + }, + { + "epoch": 29.36741214057508, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 36768 + }, + { + "epoch": 29.368210862619808, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 36769 + }, + { + "epoch": 29.369009584664536, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 36770 + }, + { + "epoch": 29.369808306709267, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 36771 + }, + { + "epoch": 29.370607028753994, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36772 + }, + { + "epoch": 29.371405750798722, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 36773 + }, + { + "epoch": 29.37220447284345, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 36774 + }, + { + "epoch": 29.373003194888177, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 36775 + }, + { + "epoch": 29.37380191693291, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 36776 + }, + { + "epoch": 29.374600638977636, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 36777 + }, + { + "epoch": 29.375399361022364, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 36778 + }, + { + "epoch": 29.37619808306709, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 36779 + }, + { + "epoch": 29.376996805111823, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 36780 + }, + { + "epoch": 29.37779552715655, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0749, + "step": 36781 + }, + { + "epoch": 29.378594249201278, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 36782 + }, + { + "epoch": 29.379392971246006, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 36783 + }, + { + "epoch": 29.380191693290733, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.1028, + "step": 36784 + }, + { + "epoch": 29.380990415335464, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 36785 + }, + { + "epoch": 29.381789137380192, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 36786 + }, + { + "epoch": 29.38258785942492, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 36787 + }, + { + "epoch": 29.383386581469647, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 36788 + }, + { + "epoch": 29.38418530351438, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 36789 + }, + { + "epoch": 29.384984025559106, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 36790 + }, + { + "epoch": 29.385782747603834, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 36791 + }, + { + "epoch": 29.38658146964856, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 36792 + }, + { + "epoch": 29.38738019169329, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 36793 + }, + { + "epoch": 29.38817891373802, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 36794 + }, + { + "epoch": 29.388977635782748, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 36795 + }, + { + "epoch": 29.389776357827476, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 36796 + }, + { + "epoch": 29.390575079872203, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 36797 + }, + { + "epoch": 29.391373801916934, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36798 + }, + { + "epoch": 29.392172523961662, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.098, + "step": 36799 + }, + { + "epoch": 29.39297124600639, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 36800 + }, + { + "epoch": 29.393769968051117, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 36801 + }, + { + "epoch": 29.394568690095845, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 36802 + }, + { + "epoch": 29.395367412140576, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 36803 + }, + { + "epoch": 29.396166134185304, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36804 + }, + { + "epoch": 29.39696485623003, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 36805 + }, + { + "epoch": 29.39776357827476, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 36806 + }, + { + "epoch": 29.39856230031949, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 36807 + }, + { + "epoch": 29.399361022364218, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 36808 + }, + { + "epoch": 29.400159744408946, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 36809 + }, + { + "epoch": 29.400958466453673, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 36810 + }, + { + "epoch": 29.4017571884984, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 36811 + }, + { + "epoch": 29.402555910543132, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 36812 + }, + { + "epoch": 29.40335463258786, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36813 + }, + { + "epoch": 29.404153354632587, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36814 + }, + { + "epoch": 29.404952076677315, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 36815 + }, + { + "epoch": 29.405750798722046, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 36816 + }, + { + "epoch": 29.406549520766774, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 36817 + }, + { + "epoch": 29.4073482428115, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 36818 + }, + { + "epoch": 29.40814696485623, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 36819 + }, + { + "epoch": 29.408945686900957, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 36820 + }, + { + "epoch": 29.409744408945688, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 36821 + }, + { + "epoch": 29.410543130990416, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 36822 + }, + { + "epoch": 29.411341853035143, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 36823 + }, + { + "epoch": 29.41214057507987, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 36824 + }, + { + "epoch": 29.412939297124602, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 36825 + }, + { + "epoch": 29.41373801916933, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 36826 + }, + { + "epoch": 29.414536741214057, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 36827 + }, + { + "epoch": 29.415335463258785, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 36828 + }, + { + "epoch": 29.416134185303516, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 36829 + }, + { + "epoch": 29.416932907348244, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 36830 + }, + { + "epoch": 29.41773162939297, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 36831 + }, + { + "epoch": 29.4185303514377, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 36832 + }, + { + "epoch": 29.419329073482427, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 36833 + }, + { + "epoch": 29.420127795527158, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 36834 + }, + { + "epoch": 29.420926517571885, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 36835 + }, + { + "epoch": 29.421725239616613, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 36836 + }, + { + "epoch": 29.42252396166134, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0751, + "step": 36837 + }, + { + "epoch": 29.423322683706072, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 36838 + }, + { + "epoch": 29.4241214057508, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 36839 + }, + { + "epoch": 29.424920127795527, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36840 + }, + { + "epoch": 29.425718849840255, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 36841 + }, + { + "epoch": 29.426517571884983, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 36842 + }, + { + "epoch": 29.427316293929714, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 36843 + }, + { + "epoch": 29.42811501597444, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 36844 + }, + { + "epoch": 29.42891373801917, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 36845 + }, + { + "epoch": 29.429712460063897, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 36846 + }, + { + "epoch": 29.430511182108628, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 36847 + }, + { + "epoch": 29.431309904153355, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 36848 + }, + { + "epoch": 29.432108626198083, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 36849 + }, + { + "epoch": 29.43290734824281, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 36850 + }, + { + "epoch": 29.43370607028754, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 36851 + }, + { + "epoch": 29.43450479233227, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 36852 + }, + { + "epoch": 29.435303514376997, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 36853 + }, + { + "epoch": 29.436102236421725, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36854 + }, + { + "epoch": 29.436900958466452, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 36855 + }, + { + "epoch": 29.437699680511184, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 36856 + }, + { + "epoch": 29.43849840255591, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 36857 + }, + { + "epoch": 29.43929712460064, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 36858 + }, + { + "epoch": 29.440095846645367, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 36859 + }, + { + "epoch": 29.440894568690094, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 36860 + }, + { + "epoch": 29.441693290734825, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 36861 + }, + { + "epoch": 29.442492012779553, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 36862 + }, + { + "epoch": 29.44329073482428, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 36863 + }, + { + "epoch": 29.44408945686901, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 36864 + }, + { + "epoch": 29.44488817891374, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 36865 + }, + { + "epoch": 29.445686900958467, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 36866 + }, + { + "epoch": 29.446485623003195, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36867 + }, + { + "epoch": 29.447284345047922, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 36868 + }, + { + "epoch": 29.44808306709265, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 36869 + }, + { + "epoch": 29.44888178913738, + "grad_norm": 0.0966796875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 36870 + }, + { + "epoch": 29.44968051118211, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.1004, + "step": 36871 + }, + { + "epoch": 29.450479233226837, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 36872 + }, + { + "epoch": 29.451277955271564, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 36873 + }, + { + "epoch": 29.452076677316295, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 36874 + }, + { + "epoch": 29.452875399361023, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 36875 + }, + { + "epoch": 29.45367412140575, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 36876 + }, + { + "epoch": 29.45447284345048, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 36877 + }, + { + "epoch": 29.455271565495206, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 36878 + }, + { + "epoch": 29.456070287539937, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 36879 + }, + { + "epoch": 29.456869009584665, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 36880 + }, + { + "epoch": 29.457667731629392, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 36881 + }, + { + "epoch": 29.45846645367412, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 36882 + }, + { + "epoch": 29.45926517571885, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 36883 + }, + { + "epoch": 29.46006389776358, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 36884 + }, + { + "epoch": 29.460862619808307, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 36885 + }, + { + "epoch": 29.461661341853034, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 36886 + }, + { + "epoch": 29.462460063897762, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 36887 + }, + { + "epoch": 29.463258785942493, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 36888 + }, + { + "epoch": 29.46405750798722, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 36889 + }, + { + "epoch": 29.46485623003195, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 36890 + }, + { + "epoch": 29.465654952076676, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 36891 + }, + { + "epoch": 29.466453674121407, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 36892 + }, + { + "epoch": 29.467252396166135, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 36893 + }, + { + "epoch": 29.468051118210862, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 36894 + }, + { + "epoch": 29.46884984025559, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 36895 + }, + { + "epoch": 29.46964856230032, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 36896 + }, + { + "epoch": 29.47044728434505, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 36897 + }, + { + "epoch": 29.471246006389777, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 36898 + }, + { + "epoch": 29.472044728434504, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 36899 + }, + { + "epoch": 29.472843450479232, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 36900 + }, + { + "epoch": 29.473642172523963, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 36901 + }, + { + "epoch": 29.47444089456869, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 36902 + }, + { + "epoch": 29.47523961661342, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 36903 + }, + { + "epoch": 29.476038338658146, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 36904 + }, + { + "epoch": 29.476837060702877, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 36905 + }, + { + "epoch": 29.477635782747605, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 36906 + }, + { + "epoch": 29.478434504792332, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 36907 + }, + { + "epoch": 29.47923322683706, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 36908 + }, + { + "epoch": 29.480031948881788, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 36909 + }, + { + "epoch": 29.48083067092652, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 36910 + }, + { + "epoch": 29.481629392971247, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 36911 + }, + { + "epoch": 29.482428115015974, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 36912 + }, + { + "epoch": 29.483226837060702, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 36913 + }, + { + "epoch": 29.484025559105433, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 36914 + }, + { + "epoch": 29.48482428115016, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 36915 + }, + { + "epoch": 29.48562300319489, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 36916 + }, + { + "epoch": 29.486421725239616, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 36917 + }, + { + "epoch": 29.487220447284344, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36918 + }, + { + "epoch": 29.488019169329075, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 36919 + }, + { + "epoch": 29.488817891373802, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 36920 + }, + { + "epoch": 29.48961661341853, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 36921 + }, + { + "epoch": 29.490415335463258, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 36922 + }, + { + "epoch": 29.49121405750799, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 36923 + }, + { + "epoch": 29.492012779552716, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 36924 + }, + { + "epoch": 29.492811501597444, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 36925 + }, + { + "epoch": 29.49361022364217, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 36926 + }, + { + "epoch": 29.4944089456869, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 36927 + }, + { + "epoch": 29.49520766773163, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 36928 + }, + { + "epoch": 29.49600638977636, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 36929 + }, + { + "epoch": 29.496805111821086, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 36930 + }, + { + "epoch": 29.497603833865814, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 36931 + }, + { + "epoch": 29.498402555910545, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 36932 + }, + { + "epoch": 29.499201277955272, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 36933 + }, + { + "epoch": 29.5, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 36934 + }, + { + "epoch": 29.500798722044728, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 36935 + }, + { + "epoch": 29.501597444089455, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 36936 + }, + { + "epoch": 29.502396166134186, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 36937 + }, + { + "epoch": 29.503194888178914, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 36938 + }, + { + "epoch": 29.50399361022364, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 36939 + }, + { + "epoch": 29.50479233226837, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 36940 + }, + { + "epoch": 29.5055910543131, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 36941 + }, + { + "epoch": 29.50638977635783, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 36942 + }, + { + "epoch": 29.507188498402556, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 36943 + }, + { + "epoch": 29.507987220447284, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 36944 + }, + { + "epoch": 29.50878594249201, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 36945 + }, + { + "epoch": 29.509584664536742, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36946 + }, + { + "epoch": 29.51038338658147, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 36947 + }, + { + "epoch": 29.511182108626198, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 36948 + }, + { + "epoch": 29.511980830670925, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 36949 + }, + { + "epoch": 29.512779552715656, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 36950 + }, + { + "epoch": 29.513578274760384, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 36951 + }, + { + "epoch": 29.51437699680511, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 36952 + }, + { + "epoch": 29.51517571884984, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 36953 + }, + { + "epoch": 29.515974440894567, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 36954 + }, + { + "epoch": 29.516773162939298, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 36955 + }, + { + "epoch": 29.517571884984026, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 36956 + }, + { + "epoch": 29.518370607028753, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 36957 + }, + { + "epoch": 29.51916932907348, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 36958 + }, + { + "epoch": 29.519968051118212, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 36959 + }, + { + "epoch": 29.52076677316294, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 36960 + }, + { + "epoch": 29.521565495207668, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36961 + }, + { + "epoch": 29.522364217252395, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 36962 + }, + { + "epoch": 29.523162939297123, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 36963 + }, + { + "epoch": 29.523961661341854, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 36964 + }, + { + "epoch": 29.52476038338658, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 36965 + }, + { + "epoch": 29.52555910543131, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 36966 + }, + { + "epoch": 29.526357827476037, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 36967 + }, + { + "epoch": 29.527156549520768, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 36968 + }, + { + "epoch": 29.527955271565496, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 36969 + }, + { + "epoch": 29.528753993610223, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 36970 + }, + { + "epoch": 29.52955271565495, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 36971 + }, + { + "epoch": 29.53035143769968, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 36972 + }, + { + "epoch": 29.53115015974441, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 36973 + }, + { + "epoch": 29.531948881789138, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 36974 + }, + { + "epoch": 29.532747603833865, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 36975 + }, + { + "epoch": 29.533546325878593, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 36976 + }, + { + "epoch": 29.534345047923324, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 36977 + }, + { + "epoch": 29.53514376996805, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 36978 + }, + { + "epoch": 29.53594249201278, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 36979 + }, + { + "epoch": 29.536741214057507, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 36980 + }, + { + "epoch": 29.537539936102235, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 36981 + }, + { + "epoch": 29.538338658146966, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 36982 + }, + { + "epoch": 29.539137380191693, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 36983 + }, + { + "epoch": 29.53993610223642, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 36984 + }, + { + "epoch": 29.54073482428115, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 36985 + }, + { + "epoch": 29.54153354632588, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 36986 + }, + { + "epoch": 29.542332268370608, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 36987 + }, + { + "epoch": 29.543130990415335, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 36988 + }, + { + "epoch": 29.543929712460063, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 36989 + }, + { + "epoch": 29.544728434504794, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 36990 + }, + { + "epoch": 29.54552715654952, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 36991 + }, + { + "epoch": 29.54632587859425, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 36992 + }, + { + "epoch": 29.547124600638977, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 36993 + }, + { + "epoch": 29.547923322683705, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 36994 + }, + { + "epoch": 29.548722044728436, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 36995 + }, + { + "epoch": 29.549520766773163, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 36996 + }, + { + "epoch": 29.55031948881789, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 36997 + }, + { + "epoch": 29.55111821086262, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 36998 + }, + { + "epoch": 29.55191693290735, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 36999 + }, + { + "epoch": 29.552715654952078, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37000 + }, + { + "epoch": 29.553514376996805, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 37001 + }, + { + "epoch": 29.554313099041533, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.1015, + "step": 37002 + }, + { + "epoch": 29.55511182108626, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 37003 + }, + { + "epoch": 29.55591054313099, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 37004 + }, + { + "epoch": 29.55670926517572, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 37005 + }, + { + "epoch": 29.557507987220447, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 37006 + }, + { + "epoch": 29.558306709265175, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0713, + "step": 37007 + }, + { + "epoch": 29.559105431309906, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 37008 + }, + { + "epoch": 29.559904153354633, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37009 + }, + { + "epoch": 29.56070287539936, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37010 + }, + { + "epoch": 29.56150159744409, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37011 + }, + { + "epoch": 29.562300319488816, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 37012 + }, + { + "epoch": 29.563099041533548, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 37013 + }, + { + "epoch": 29.563897763578275, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 37014 + }, + { + "epoch": 29.564696485623003, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 37015 + }, + { + "epoch": 29.56549520766773, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37016 + }, + { + "epoch": 29.56629392971246, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 37017 + }, + { + "epoch": 29.56709265175719, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37018 + }, + { + "epoch": 29.567891373801917, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 37019 + }, + { + "epoch": 29.568690095846645, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 37020 + }, + { + "epoch": 29.569488817891372, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 37021 + }, + { + "epoch": 29.570287539936103, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 37022 + }, + { + "epoch": 29.57108626198083, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 37023 + }, + { + "epoch": 29.57188498402556, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 37024 + }, + { + "epoch": 29.572683706070286, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 37025 + }, + { + "epoch": 29.573482428115017, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 37026 + }, + { + "epoch": 29.574281150159745, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 37027 + }, + { + "epoch": 29.575079872204473, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 37028 + }, + { + "epoch": 29.5758785942492, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 37029 + }, + { + "epoch": 29.576677316293928, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 37030 + }, + { + "epoch": 29.57747603833866, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 37031 + }, + { + "epoch": 29.578274760383387, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 37032 + }, + { + "epoch": 29.579073482428115, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 37033 + }, + { + "epoch": 29.579872204472842, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37034 + }, + { + "epoch": 29.580670926517573, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 37035 + }, + { + "epoch": 29.5814696485623, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37036 + }, + { + "epoch": 29.58226837060703, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 37037 + }, + { + "epoch": 29.583067092651756, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 37038 + }, + { + "epoch": 29.583865814696484, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 37039 + }, + { + "epoch": 29.584664536741215, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 37040 + }, + { + "epoch": 29.585463258785943, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 37041 + }, + { + "epoch": 29.58626198083067, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 37042 + }, + { + "epoch": 29.587060702875398, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 37043 + }, + { + "epoch": 29.58785942492013, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 37044 + }, + { + "epoch": 29.588658146964857, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 37045 + }, + { + "epoch": 29.589456869009584, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 37046 + }, + { + "epoch": 29.590255591054312, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 37047 + }, + { + "epoch": 29.591054313099043, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 37048 + }, + { + "epoch": 29.59185303514377, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 37049 + }, + { + "epoch": 29.5926517571885, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 37050 + }, + { + "epoch": 29.593450479233226, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 37051 + }, + { + "epoch": 29.594249201277954, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 37052 + }, + { + "epoch": 29.595047923322685, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 37053 + }, + { + "epoch": 29.595846645367413, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 37054 + }, + { + "epoch": 29.59664536741214, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 37055 + }, + { + "epoch": 29.597444089456868, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 37056 + }, + { + "epoch": 29.5982428115016, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 37057 + }, + { + "epoch": 29.599041533546327, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 37058 + }, + { + "epoch": 29.599840255591054, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 37059 + }, + { + "epoch": 29.600638977635782, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 37060 + }, + { + "epoch": 29.60143769968051, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 37061 + }, + { + "epoch": 29.60223642172524, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 37062 + }, + { + "epoch": 29.60303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 37063 + }, + { + "epoch": 29.603833865814696, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37064 + }, + { + "epoch": 29.604632587859424, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 37065 + }, + { + "epoch": 29.605431309904155, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 37066 + }, + { + "epoch": 29.606230031948883, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 37067 + }, + { + "epoch": 29.60702875399361, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 37068 + }, + { + "epoch": 29.607827476038338, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 37069 + }, + { + "epoch": 29.608626198083066, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0738, + "step": 37070 + }, + { + "epoch": 29.609424920127797, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 37071 + }, + { + "epoch": 29.610223642172524, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 37072 + }, + { + "epoch": 29.611022364217252, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 37073 + }, + { + "epoch": 29.61182108626198, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 37074 + }, + { + "epoch": 29.61261980830671, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 37075 + }, + { + "epoch": 29.61341853035144, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 37076 + }, + { + "epoch": 29.614217252396166, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 37077 + }, + { + "epoch": 29.615015974440894, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 37078 + }, + { + "epoch": 29.61581469648562, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 37079 + }, + { + "epoch": 29.616613418530353, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 37080 + }, + { + "epoch": 29.61741214057508, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 37081 + }, + { + "epoch": 29.618210862619808, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 37082 + }, + { + "epoch": 29.619009584664536, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 37083 + }, + { + "epoch": 29.619808306709267, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 37084 + }, + { + "epoch": 29.620607028753994, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 37085 + }, + { + "epoch": 29.621405750798722, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 37086 + }, + { + "epoch": 29.62220447284345, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 37087 + }, + { + "epoch": 29.623003194888177, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37088 + }, + { + "epoch": 29.62380191693291, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 37089 + }, + { + "epoch": 29.624600638977636, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 37090 + }, + { + "epoch": 29.625399361022364, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 37091 + }, + { + "epoch": 29.62619808306709, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 37092 + }, + { + "epoch": 29.626996805111823, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 37093 + }, + { + "epoch": 29.62779552715655, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 37094 + }, + { + "epoch": 29.628594249201278, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 37095 + }, + { + "epoch": 29.629392971246006, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 37096 + }, + { + "epoch": 29.630191693290733, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 37097 + }, + { + "epoch": 29.630990415335464, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37098 + }, + { + "epoch": 29.631789137380192, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 37099 + }, + { + "epoch": 29.63258785942492, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 37100 + }, + { + "epoch": 29.633386581469647, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 37101 + }, + { + "epoch": 29.63418530351438, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 37102 + }, + { + "epoch": 29.634984025559106, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 37103 + }, + { + "epoch": 29.635782747603834, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37104 + }, + { + "epoch": 29.63658146964856, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 37105 + }, + { + "epoch": 29.63738019169329, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.1002, + "step": 37106 + }, + { + "epoch": 29.63817891373802, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 37107 + }, + { + "epoch": 29.638977635782748, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 37108 + }, + { + "epoch": 29.639776357827476, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 37109 + }, + { + "epoch": 29.640575079872203, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 37110 + }, + { + "epoch": 29.641373801916934, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 37111 + }, + { + "epoch": 29.642172523961662, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37112 + }, + { + "epoch": 29.64297124600639, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 37113 + }, + { + "epoch": 29.643769968051117, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37114 + }, + { + "epoch": 29.644568690095845, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 37115 + }, + { + "epoch": 29.645367412140576, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 37116 + }, + { + "epoch": 29.646166134185304, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 37117 + }, + { + "epoch": 29.64696485623003, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 37118 + }, + { + "epoch": 29.64776357827476, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 37119 + }, + { + "epoch": 29.64856230031949, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 37120 + }, + { + "epoch": 29.649361022364218, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 37121 + }, + { + "epoch": 29.650159744408946, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 37122 + }, + { + "epoch": 29.650958466453673, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 37123 + }, + { + "epoch": 29.6517571884984, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 37124 + }, + { + "epoch": 29.652555910543132, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0745, + "step": 37125 + }, + { + "epoch": 29.65335463258786, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37126 + }, + { + "epoch": 29.654153354632587, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 37127 + }, + { + "epoch": 29.654952076677315, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 37128 + }, + { + "epoch": 29.655750798722046, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 37129 + }, + { + "epoch": 29.656549520766774, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 37130 + }, + { + "epoch": 29.6573482428115, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 37131 + }, + { + "epoch": 29.65814696485623, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 37132 + }, + { + "epoch": 29.658945686900957, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37133 + }, + { + "epoch": 29.659744408945688, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 37134 + }, + { + "epoch": 29.660543130990416, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 37135 + }, + { + "epoch": 29.661341853035143, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 37136 + }, + { + "epoch": 29.66214057507987, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 37137 + }, + { + "epoch": 29.662939297124602, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 37138 + }, + { + "epoch": 29.66373801916933, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 37139 + }, + { + "epoch": 29.664536741214057, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 37140 + }, + { + "epoch": 29.665335463258785, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 37141 + }, + { + "epoch": 29.666134185303516, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 37142 + }, + { + "epoch": 29.666932907348244, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 37143 + }, + { + "epoch": 29.66773162939297, + "grad_norm": 0.25, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 37144 + }, + { + "epoch": 29.6685303514377, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 37145 + }, + { + "epoch": 29.669329073482427, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 37146 + }, + { + "epoch": 29.670127795527158, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 37147 + }, + { + "epoch": 29.670926517571885, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 37148 + }, + { + "epoch": 29.671725239616613, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 37149 + }, + { + "epoch": 29.67252396166134, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 37150 + }, + { + "epoch": 29.673322683706072, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 37151 + }, + { + "epoch": 29.6741214057508, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37152 + }, + { + "epoch": 29.674920127795527, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 37153 + }, + { + "epoch": 29.675718849840255, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 37154 + }, + { + "epoch": 29.676517571884983, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 37155 + }, + { + "epoch": 29.677316293929714, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37156 + }, + { + "epoch": 29.67811501597444, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 37157 + }, + { + "epoch": 29.67891373801917, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 37158 + }, + { + "epoch": 29.679712460063897, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 37159 + }, + { + "epoch": 29.680511182108628, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 37160 + }, + { + "epoch": 29.681309904153355, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 37161 + }, + { + "epoch": 29.682108626198083, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 37162 + }, + { + "epoch": 29.68290734824281, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 37163 + }, + { + "epoch": 29.68370607028754, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 37164 + }, + { + "epoch": 29.68450479233227, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 37165 + }, + { + "epoch": 29.685303514376997, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 37166 + }, + { + "epoch": 29.686102236421725, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 37167 + }, + { + "epoch": 29.686900958466452, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 37168 + }, + { + "epoch": 29.687699680511184, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 37169 + }, + { + "epoch": 29.68849840255591, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 37170 + }, + { + "epoch": 29.68929712460064, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37171 + }, + { + "epoch": 29.690095846645367, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 37172 + }, + { + "epoch": 29.690894568690094, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 37173 + }, + { + "epoch": 29.691693290734825, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37174 + }, + { + "epoch": 29.692492012779553, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 37175 + }, + { + "epoch": 29.69329073482428, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 37176 + }, + { + "epoch": 29.69408945686901, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 37177 + }, + { + "epoch": 29.69488817891374, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 37178 + }, + { + "epoch": 29.695686900958467, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 37179 + }, + { + "epoch": 29.696485623003195, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 37180 + }, + { + "epoch": 29.697284345047922, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37181 + }, + { + "epoch": 29.69808306709265, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37182 + }, + { + "epoch": 29.69888178913738, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 37183 + }, + { + "epoch": 29.69968051118211, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 37184 + }, + { + "epoch": 29.700479233226837, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 37185 + }, + { + "epoch": 29.701277955271564, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 37186 + }, + { + "epoch": 29.702076677316295, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 37187 + }, + { + "epoch": 29.702875399361023, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 37188 + }, + { + "epoch": 29.70367412140575, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 37189 + }, + { + "epoch": 29.70447284345048, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 37190 + }, + { + "epoch": 29.705271565495206, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37191 + }, + { + "epoch": 29.706070287539937, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 37192 + }, + { + "epoch": 29.706869009584665, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 37193 + }, + { + "epoch": 29.707667731629392, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 37194 + }, + { + "epoch": 29.70846645367412, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 37195 + }, + { + "epoch": 29.70926517571885, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0725, + "step": 37196 + }, + { + "epoch": 29.71006389776358, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 37197 + }, + { + "epoch": 29.710862619808307, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 37198 + }, + { + "epoch": 29.711661341853034, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 37199 + }, + { + "epoch": 29.712460063897765, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 37200 + }, + { + "epoch": 29.713258785942493, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 37201 + }, + { + "epoch": 29.71405750798722, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 37202 + }, + { + "epoch": 29.71485623003195, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 37203 + }, + { + "epoch": 29.715654952076676, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 37204 + }, + { + "epoch": 29.716453674121407, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37205 + }, + { + "epoch": 29.717252396166135, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 37206 + }, + { + "epoch": 29.718051118210862, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37207 + }, + { + "epoch": 29.71884984025559, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 37208 + }, + { + "epoch": 29.71964856230032, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 37209 + }, + { + "epoch": 29.72044728434505, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 37210 + }, + { + "epoch": 29.721246006389777, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 37211 + }, + { + "epoch": 29.722044728434504, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37212 + }, + { + "epoch": 29.722843450479232, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 37213 + }, + { + "epoch": 29.723642172523963, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 37214 + }, + { + "epoch": 29.72444089456869, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 37215 + }, + { + "epoch": 29.72523961661342, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37216 + }, + { + "epoch": 29.726038338658146, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 37217 + }, + { + "epoch": 29.726837060702877, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 37218 + }, + { + "epoch": 29.727635782747605, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 37219 + }, + { + "epoch": 29.728434504792332, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 37220 + }, + { + "epoch": 29.72923322683706, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 37221 + }, + { + "epoch": 29.730031948881788, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 37222 + }, + { + "epoch": 29.73083067092652, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 37223 + }, + { + "epoch": 29.731629392971247, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 37224 + }, + { + "epoch": 29.732428115015974, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37225 + }, + { + "epoch": 29.733226837060702, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 37226 + }, + { + "epoch": 29.734025559105433, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 37227 + }, + { + "epoch": 29.73482428115016, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 37228 + }, + { + "epoch": 29.73562300319489, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 37229 + }, + { + "epoch": 29.736421725239616, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 37230 + }, + { + "epoch": 29.737220447284344, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 37231 + }, + { + "epoch": 29.738019169329075, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 37232 + }, + { + "epoch": 29.738817891373802, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 37233 + }, + { + "epoch": 29.73961661341853, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 37234 + }, + { + "epoch": 29.740415335463258, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 37235 + }, + { + "epoch": 29.74121405750799, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 37236 + }, + { + "epoch": 29.742012779552716, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 37237 + }, + { + "epoch": 29.742811501597444, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37238 + }, + { + "epoch": 29.74361022364217, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 37239 + }, + { + "epoch": 29.7444089456869, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 37240 + }, + { + "epoch": 29.74520766773163, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 37241 + }, + { + "epoch": 29.74600638977636, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 37242 + }, + { + "epoch": 29.746805111821086, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0735, + "step": 37243 + }, + { + "epoch": 29.747603833865814, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 37244 + }, + { + "epoch": 29.748402555910545, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 37245 + }, + { + "epoch": 29.749201277955272, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37246 + }, + { + "epoch": 29.75, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 37247 + }, + { + "epoch": 29.750798722044728, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 37248 + }, + { + "epoch": 29.751597444089455, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37249 + }, + { + "epoch": 29.752396166134186, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37250 + }, + { + "epoch": 29.753194888178914, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 37251 + }, + { + "epoch": 29.75399361022364, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 37252 + }, + { + "epoch": 29.75479233226837, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 37253 + }, + { + "epoch": 29.7555910543131, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37254 + }, + { + "epoch": 29.75638977635783, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 37255 + }, + { + "epoch": 29.757188498402556, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37256 + }, + { + "epoch": 29.757987220447284, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 37257 + }, + { + "epoch": 29.75878594249201, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 37258 + }, + { + "epoch": 29.759584664536742, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0983, + "step": 37259 + }, + { + "epoch": 29.76038338658147, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 37260 + }, + { + "epoch": 29.761182108626198, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 37261 + }, + { + "epoch": 29.761980830670925, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 37262 + }, + { + "epoch": 29.762779552715656, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 37263 + }, + { + "epoch": 29.763578274760384, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 37264 + }, + { + "epoch": 29.76437699680511, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 37265 + }, + { + "epoch": 29.76517571884984, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 37266 + }, + { + "epoch": 29.765974440894567, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 37267 + }, + { + "epoch": 29.766773162939298, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 37268 + }, + { + "epoch": 29.767571884984026, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 37269 + }, + { + "epoch": 29.768370607028753, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 37270 + }, + { + "epoch": 29.76916932907348, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 37271 + }, + { + "epoch": 29.769968051118212, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 37272 + }, + { + "epoch": 29.77076677316294, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 37273 + }, + { + "epoch": 29.771565495207668, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 37274 + }, + { + "epoch": 29.772364217252395, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 37275 + }, + { + "epoch": 29.773162939297123, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37276 + }, + { + "epoch": 29.773961661341854, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 37277 + }, + { + "epoch": 29.77476038338658, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 37278 + }, + { + "epoch": 29.77555910543131, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 37279 + }, + { + "epoch": 29.776357827476037, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37280 + }, + { + "epoch": 29.777156549520768, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 37281 + }, + { + "epoch": 29.777955271565496, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37282 + }, + { + "epoch": 29.778753993610223, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 37283 + }, + { + "epoch": 29.77955271565495, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 37284 + }, + { + "epoch": 29.78035143769968, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 37285 + }, + { + "epoch": 29.78115015974441, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 37286 + }, + { + "epoch": 29.781948881789138, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 37287 + }, + { + "epoch": 29.782747603833865, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 37288 + }, + { + "epoch": 29.783546325878593, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 37289 + }, + { + "epoch": 29.784345047923324, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 37290 + }, + { + "epoch": 29.78514376996805, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 37291 + }, + { + "epoch": 29.78594249201278, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 37292 + }, + { + "epoch": 29.786741214057507, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 37293 + }, + { + "epoch": 29.787539936102235, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 37294 + }, + { + "epoch": 29.788338658146966, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37295 + }, + { + "epoch": 29.789137380191693, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 37296 + }, + { + "epoch": 29.78993610223642, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0755, + "step": 37297 + }, + { + "epoch": 29.79073482428115, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 37298 + }, + { + "epoch": 29.79153354632588, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 37299 + }, + { + "epoch": 29.792332268370608, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 37300 + }, + { + "epoch": 29.793130990415335, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37301 + }, + { + "epoch": 29.793929712460063, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 37302 + }, + { + "epoch": 29.794728434504794, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 37303 + }, + { + "epoch": 29.79552715654952, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 37304 + }, + { + "epoch": 29.79632587859425, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 37305 + }, + { + "epoch": 29.797124600638977, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 37306 + }, + { + "epoch": 29.797923322683705, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 37307 + }, + { + "epoch": 29.798722044728436, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 37308 + }, + { + "epoch": 29.799520766773163, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 37309 + }, + { + "epoch": 29.80031948881789, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 37310 + }, + { + "epoch": 29.80111821086262, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 37311 + }, + { + "epoch": 29.80191693290735, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 37312 + }, + { + "epoch": 29.802715654952078, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 37313 + }, + { + "epoch": 29.803514376996805, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 37314 + }, + { + "epoch": 29.804313099041533, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 37315 + }, + { + "epoch": 29.80511182108626, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.1021, + "step": 37316 + }, + { + "epoch": 29.80591054313099, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 37317 + }, + { + "epoch": 29.80670926517572, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 37318 + }, + { + "epoch": 29.807507987220447, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 37319 + }, + { + "epoch": 29.808306709265175, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 37320 + }, + { + "epoch": 29.809105431309906, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 37321 + }, + { + "epoch": 29.809904153354633, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37322 + }, + { + "epoch": 29.81070287539936, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 37323 + }, + { + "epoch": 29.81150159744409, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 37324 + }, + { + "epoch": 29.812300319488816, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37325 + }, + { + "epoch": 29.813099041533548, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 37326 + }, + { + "epoch": 29.813897763578275, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 37327 + }, + { + "epoch": 29.814696485623003, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 37328 + }, + { + "epoch": 29.81549520766773, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 37329 + }, + { + "epoch": 29.81629392971246, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 37330 + }, + { + "epoch": 29.81709265175719, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37331 + }, + { + "epoch": 29.817891373801917, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 37332 + }, + { + "epoch": 29.818690095846645, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 37333 + }, + { + "epoch": 29.819488817891372, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 37334 + }, + { + "epoch": 29.820287539936103, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 37335 + }, + { + "epoch": 29.82108626198083, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 37336 + }, + { + "epoch": 29.82188498402556, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 37337 + }, + { + "epoch": 29.822683706070286, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 37338 + }, + { + "epoch": 29.823482428115017, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 37339 + }, + { + "epoch": 29.824281150159745, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 37340 + }, + { + "epoch": 29.825079872204473, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 37341 + }, + { + "epoch": 29.8258785942492, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37342 + }, + { + "epoch": 29.826677316293928, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 37343 + }, + { + "epoch": 29.82747603833866, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 37344 + }, + { + "epoch": 29.828274760383387, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 37345 + }, + { + "epoch": 29.829073482428115, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 37346 + }, + { + "epoch": 29.829872204472842, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 37347 + }, + { + "epoch": 29.830670926517573, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0753, + "step": 37348 + }, + { + "epoch": 29.8314696485623, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 37349 + }, + { + "epoch": 29.83226837060703, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 37350 + }, + { + "epoch": 29.833067092651756, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37351 + }, + { + "epoch": 29.833865814696484, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 37352 + }, + { + "epoch": 29.834664536741215, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 37353 + }, + { + "epoch": 29.835463258785943, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 37354 + }, + { + "epoch": 29.83626198083067, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 37355 + }, + { + "epoch": 29.837060702875398, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 37356 + }, + { + "epoch": 29.83785942492013, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 37357 + }, + { + "epoch": 29.838658146964857, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 37358 + }, + { + "epoch": 29.839456869009584, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37359 + }, + { + "epoch": 29.840255591054312, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 37360 + }, + { + "epoch": 29.841054313099043, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 37361 + }, + { + "epoch": 29.84185303514377, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 37362 + }, + { + "epoch": 29.8426517571885, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 37363 + }, + { + "epoch": 29.843450479233226, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 37364 + }, + { + "epoch": 29.844249201277954, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 37365 + }, + { + "epoch": 29.845047923322685, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 37366 + }, + { + "epoch": 29.845846645367413, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 37367 + }, + { + "epoch": 29.84664536741214, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0753, + "step": 37368 + }, + { + "epoch": 29.847444089456868, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 37369 + }, + { + "epoch": 29.8482428115016, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0982, + "step": 37370 + }, + { + "epoch": 29.849041533546327, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 37371 + }, + { + "epoch": 29.849840255591054, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 37372 + }, + { + "epoch": 29.850638977635782, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 37373 + }, + { + "epoch": 29.85143769968051, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 37374 + }, + { + "epoch": 29.85223642172524, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 37375 + }, + { + "epoch": 29.85303514376997, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 37376 + }, + { + "epoch": 29.853833865814696, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37377 + }, + { + "epoch": 29.854632587859424, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37378 + }, + { + "epoch": 29.855431309904155, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 37379 + }, + { + "epoch": 29.856230031948883, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 37380 + }, + { + "epoch": 29.85702875399361, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 37381 + }, + { + "epoch": 29.857827476038338, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 37382 + }, + { + "epoch": 29.858626198083066, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 37383 + }, + { + "epoch": 29.859424920127797, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 37384 + }, + { + "epoch": 29.860223642172524, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 37385 + }, + { + "epoch": 29.861022364217252, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 37386 + }, + { + "epoch": 29.86182108626198, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 37387 + }, + { + "epoch": 29.86261980830671, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 37388 + }, + { + "epoch": 29.86341853035144, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 37389 + }, + { + "epoch": 29.864217252396166, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 37390 + }, + { + "epoch": 29.865015974440894, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 37391 + }, + { + "epoch": 29.86581469648562, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 37392 + }, + { + "epoch": 29.866613418530353, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 37393 + }, + { + "epoch": 29.86741214057508, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 37394 + }, + { + "epoch": 29.868210862619808, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37395 + }, + { + "epoch": 29.869009584664536, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 37396 + }, + { + "epoch": 29.869808306709267, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 37397 + }, + { + "epoch": 29.870607028753994, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 37398 + }, + { + "epoch": 29.871405750798722, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 37399 + }, + { + "epoch": 29.87220447284345, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37400 + }, + { + "epoch": 29.873003194888177, + "grad_norm": 0.2373046875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37401 + }, + { + "epoch": 29.87380191693291, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37402 + }, + { + "epoch": 29.874600638977636, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 37403 + }, + { + "epoch": 29.875399361022364, + "grad_norm": 0.294921875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 37404 + }, + { + "epoch": 29.87619808306709, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 37405 + }, + { + "epoch": 29.876996805111823, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37406 + }, + { + "epoch": 29.87779552715655, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 37407 + }, + { + "epoch": 29.878594249201278, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 37408 + }, + { + "epoch": 29.879392971246006, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 37409 + }, + { + "epoch": 29.880191693290733, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 37410 + }, + { + "epoch": 29.880990415335464, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 37411 + }, + { + "epoch": 29.881789137380192, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 37412 + }, + { + "epoch": 29.88258785942492, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 37413 + }, + { + "epoch": 29.883386581469647, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 37414 + }, + { + "epoch": 29.88418530351438, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 37415 + }, + { + "epoch": 29.884984025559106, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 37416 + }, + { + "epoch": 29.885782747603834, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 37417 + }, + { + "epoch": 29.88658146964856, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 37418 + }, + { + "epoch": 29.88738019169329, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 37419 + }, + { + "epoch": 29.88817891373802, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 37420 + }, + { + "epoch": 29.888977635782748, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 37421 + }, + { + "epoch": 29.889776357827476, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 37422 + }, + { + "epoch": 29.890575079872203, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 37423 + }, + { + "epoch": 29.891373801916934, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 37424 + }, + { + "epoch": 29.892172523961662, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 37425 + }, + { + "epoch": 29.89297124600639, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 37426 + }, + { + "epoch": 29.893769968051117, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 37427 + }, + { + "epoch": 29.894568690095845, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 37428 + }, + { + "epoch": 29.895367412140576, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 37429 + }, + { + "epoch": 29.896166134185304, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 37430 + }, + { + "epoch": 29.89696485623003, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37431 + }, + { + "epoch": 29.89776357827476, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 37432 + }, + { + "epoch": 29.89856230031949, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 37433 + }, + { + "epoch": 29.899361022364218, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 37434 + }, + { + "epoch": 29.900159744408946, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 37435 + }, + { + "epoch": 29.900958466453673, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 37436 + }, + { + "epoch": 29.9017571884984, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 37437 + }, + { + "epoch": 29.902555910543132, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 37438 + }, + { + "epoch": 29.90335463258786, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 37439 + }, + { + "epoch": 29.904153354632587, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 37440 + }, + { + "epoch": 29.904952076677315, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37441 + }, + { + "epoch": 29.905750798722046, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 37442 + }, + { + "epoch": 29.906549520766774, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 37443 + }, + { + "epoch": 29.9073482428115, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 37444 + }, + { + "epoch": 29.90814696485623, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 37445 + }, + { + "epoch": 29.908945686900957, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 37446 + }, + { + "epoch": 29.909744408945688, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 37447 + }, + { + "epoch": 29.910543130990416, + "grad_norm": 0.09130859375, + "learning_rate": 0.0001, + "loss": 1.0973, + "step": 37448 + }, + { + "epoch": 29.911341853035143, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 37449 + }, + { + "epoch": 29.91214057507987, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 37450 + }, + { + "epoch": 29.912939297124602, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 37451 + }, + { + "epoch": 29.91373801916933, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 37452 + }, + { + "epoch": 29.914536741214057, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 37453 + }, + { + "epoch": 29.915335463258785, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 37454 + }, + { + "epoch": 29.916134185303516, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 37455 + }, + { + "epoch": 29.916932907348244, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 37456 + }, + { + "epoch": 29.91773162939297, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 37457 + }, + { + "epoch": 29.9185303514377, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 37458 + }, + { + "epoch": 29.919329073482427, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 37459 + }, + { + "epoch": 29.920127795527158, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 37460 + }, + { + "epoch": 29.920926517571885, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 37461 + }, + { + "epoch": 29.921725239616613, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 37462 + }, + { + "epoch": 29.92252396166134, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 37463 + }, + { + "epoch": 29.923322683706072, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 37464 + }, + { + "epoch": 29.9241214057508, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 37465 + }, + { + "epoch": 29.924920127795527, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 37466 + }, + { + "epoch": 29.925718849840255, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 37467 + }, + { + "epoch": 29.926517571884983, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 37468 + }, + { + "epoch": 29.927316293929714, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 37469 + }, + { + "epoch": 29.92811501597444, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 37470 + }, + { + "epoch": 29.92891373801917, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 37471 + }, + { + "epoch": 29.929712460063897, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37472 + }, + { + "epoch": 29.930511182108628, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37473 + }, + { + "epoch": 29.931309904153355, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 37474 + }, + { + "epoch": 29.932108626198083, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 37475 + }, + { + "epoch": 29.93290734824281, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 37476 + }, + { + "epoch": 29.93370607028754, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 37477 + }, + { + "epoch": 29.93450479233227, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 37478 + }, + { + "epoch": 29.935303514376997, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 37479 + }, + { + "epoch": 29.936102236421725, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37480 + }, + { + "epoch": 29.936900958466452, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 37481 + }, + { + "epoch": 29.937699680511184, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 37482 + }, + { + "epoch": 29.93849840255591, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 37483 + }, + { + "epoch": 29.93929712460064, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 37484 + }, + { + "epoch": 29.940095846645367, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 37485 + }, + { + "epoch": 29.940894568690094, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 37486 + }, + { + "epoch": 29.941693290734825, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 37487 + }, + { + "epoch": 29.942492012779553, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 37488 + }, + { + "epoch": 29.94329073482428, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 37489 + }, + { + "epoch": 29.94408945686901, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 37490 + }, + { + "epoch": 29.94488817891374, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 37491 + }, + { + "epoch": 29.945686900958467, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 37492 + }, + { + "epoch": 29.946485623003195, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37493 + }, + { + "epoch": 29.947284345047922, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 37494 + }, + { + "epoch": 29.94808306709265, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 37495 + }, + { + "epoch": 29.94888178913738, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0761, + "step": 37496 + }, + { + "epoch": 29.94968051118211, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37497 + }, + { + "epoch": 29.950479233226837, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 37498 + }, + { + "epoch": 29.951277955271564, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 37499 + }, + { + "epoch": 29.952076677316295, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37500 + }, + { + "epoch": 29.952875399361023, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 37501 + }, + { + "epoch": 29.95367412140575, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 37502 + }, + { + "epoch": 29.95447284345048, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 37503 + }, + { + "epoch": 29.955271565495206, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 37504 + }, + { + "epoch": 29.956070287539937, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 37505 + }, + { + "epoch": 29.956869009584665, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 37506 + }, + { + "epoch": 29.957667731629392, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 37507 + }, + { + "epoch": 29.95846645367412, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 37508 + }, + { + "epoch": 29.95926517571885, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0757, + "step": 37509 + }, + { + "epoch": 29.96006389776358, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 37510 + }, + { + "epoch": 29.960862619808307, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 37511 + }, + { + "epoch": 29.961661341853034, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 37512 + }, + { + "epoch": 29.962460063897765, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 37513 + }, + { + "epoch": 29.963258785942493, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 37514 + }, + { + "epoch": 29.96405750798722, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 37515 + }, + { + "epoch": 29.96485623003195, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 37516 + }, + { + "epoch": 29.965654952076676, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 37517 + }, + { + "epoch": 29.966453674121407, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 37518 + }, + { + "epoch": 29.967252396166135, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 37519 + }, + { + "epoch": 29.968051118210862, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0761, + "step": 37520 + }, + { + "epoch": 29.96884984025559, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 37521 + }, + { + "epoch": 29.96964856230032, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 37522 + }, + { + "epoch": 29.97044728434505, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 37523 + }, + { + "epoch": 29.971246006389777, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 37524 + }, + { + "epoch": 29.972044728434504, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 37525 + }, + { + "epoch": 29.972843450479232, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 37526 + }, + { + "epoch": 29.973642172523963, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 37527 + }, + { + "epoch": 29.97444089456869, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 37528 + }, + { + "epoch": 29.97523961661342, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 37529 + }, + { + "epoch": 29.976038338658146, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 37530 + }, + { + "epoch": 29.976837060702877, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 37531 + }, + { + "epoch": 29.977635782747605, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 37532 + }, + { + "epoch": 29.978434504792332, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 37533 + }, + { + "epoch": 29.97923322683706, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 37534 + }, + { + "epoch": 29.980031948881788, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 37535 + }, + { + "epoch": 29.98083067092652, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 37536 + }, + { + "epoch": 29.981629392971247, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 37537 + }, + { + "epoch": 29.982428115015974, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 37538 + }, + { + "epoch": 29.983226837060702, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 37539 + }, + { + "epoch": 29.984025559105433, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 37540 + }, + { + "epoch": 29.98482428115016, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 37541 + }, + { + "epoch": 29.98562300319489, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 37542 + }, + { + "epoch": 29.986421725239616, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 37543 + }, + { + "epoch": 29.987220447284344, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 37544 + }, + { + "epoch": 29.988019169329075, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 37545 + }, + { + "epoch": 29.988817891373802, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 37546 + }, + { + "epoch": 29.98961661341853, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37547 + }, + { + "epoch": 29.990415335463258, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 37548 + }, + { + "epoch": 29.99121405750799, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37549 + }, + { + "epoch": 29.992012779552716, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 37550 + }, + { + "epoch": 29.992811501597444, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 37551 + }, + { + "epoch": 29.99361022364217, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 37552 + }, + { + "epoch": 29.9944089456869, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 37553 + }, + { + "epoch": 29.99520766773163, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 37554 + }, + { + "epoch": 29.99600638977636, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 37555 + }, + { + "epoch": 29.996805111821086, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 37556 + }, + { + "epoch": 29.997603833865814, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 37557 + }, + { + "epoch": 29.998402555910545, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 37558 + }, + { + "epoch": 29.999201277955272, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37559 + }, + { + "epoch": 30.0, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37560 + }, + { + "epoch": 30.000798722044728, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 37561 + }, + { + "epoch": 30.001597444089455, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 37562 + }, + { + "epoch": 30.002396166134186, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 37563 + }, + { + "epoch": 30.003194888178914, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37564 + }, + { + "epoch": 30.00399361022364, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 37565 + }, + { + "epoch": 30.00479233226837, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 37566 + }, + { + "epoch": 30.0055910543131, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 37567 + }, + { + "epoch": 30.00638977635783, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37568 + }, + { + "epoch": 30.007188498402556, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 37569 + }, + { + "epoch": 30.007987220447284, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 37570 + }, + { + "epoch": 30.00878594249201, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 37571 + }, + { + "epoch": 30.009584664536742, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 37572 + }, + { + "epoch": 30.01038338658147, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 37573 + }, + { + "epoch": 30.011182108626198, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 37574 + }, + { + "epoch": 30.011980830670925, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 37575 + }, + { + "epoch": 30.012779552715656, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 37576 + }, + { + "epoch": 30.013578274760384, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 37577 + }, + { + "epoch": 30.01437699680511, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37578 + }, + { + "epoch": 30.01517571884984, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 37579 + }, + { + "epoch": 30.015974440894567, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 37580 + }, + { + "epoch": 30.016773162939298, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 37581 + }, + { + "epoch": 30.017571884984026, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0756, + "step": 37582 + }, + { + "epoch": 30.018370607028753, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 37583 + }, + { + "epoch": 30.01916932907348, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 37584 + }, + { + "epoch": 30.019968051118212, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 37585 + }, + { + "epoch": 30.02076677316294, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 37586 + }, + { + "epoch": 30.021565495207668, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 37587 + }, + { + "epoch": 30.022364217252395, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 37588 + }, + { + "epoch": 30.023162939297123, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 37589 + }, + { + "epoch": 30.023961661341854, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 37590 + }, + { + "epoch": 30.02476038338658, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 37591 + }, + { + "epoch": 30.02555910543131, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37592 + }, + { + "epoch": 30.026357827476037, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 37593 + }, + { + "epoch": 30.027156549520768, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 37594 + }, + { + "epoch": 30.027955271565496, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37595 + }, + { + "epoch": 30.028753993610223, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0726, + "step": 37596 + }, + { + "epoch": 30.02955271565495, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0756, + "step": 37597 + }, + { + "epoch": 30.03035143769968, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 37598 + }, + { + "epoch": 30.03115015974441, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 37599 + }, + { + "epoch": 30.031948881789138, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 37600 + }, + { + "epoch": 30.032747603833865, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 37601 + }, + { + "epoch": 30.033546325878593, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 37602 + }, + { + "epoch": 30.034345047923324, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0942, + "step": 37603 + }, + { + "epoch": 30.03514376996805, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 37604 + }, + { + "epoch": 30.03594249201278, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 37605 + }, + { + "epoch": 30.036741214057507, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 37606 + }, + { + "epoch": 30.037539936102238, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 37607 + }, + { + "epoch": 30.038338658146966, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 37608 + }, + { + "epoch": 30.039137380191693, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 37609 + }, + { + "epoch": 30.03993610223642, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 37610 + }, + { + "epoch": 30.04073482428115, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 37611 + }, + { + "epoch": 30.04153354632588, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 37612 + }, + { + "epoch": 30.042332268370608, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 37613 + }, + { + "epoch": 30.043130990415335, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 37614 + }, + { + "epoch": 30.043929712460063, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 37615 + }, + { + "epoch": 30.044728434504794, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 37616 + }, + { + "epoch": 30.04552715654952, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 37617 + }, + { + "epoch": 30.04632587859425, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 37618 + }, + { + "epoch": 30.047124600638977, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 37619 + }, + { + "epoch": 30.047923322683705, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37620 + }, + { + "epoch": 30.048722044728436, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 37621 + }, + { + "epoch": 30.049520766773163, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 37622 + }, + { + "epoch": 30.05031948881789, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 37623 + }, + { + "epoch": 30.05111821086262, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37624 + }, + { + "epoch": 30.05191693290735, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 37625 + }, + { + "epoch": 30.052715654952078, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 37626 + }, + { + "epoch": 30.053514376996805, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37627 + }, + { + "epoch": 30.054313099041533, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 37628 + }, + { + "epoch": 30.05511182108626, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 37629 + }, + { + "epoch": 30.05591054313099, + "grad_norm": 0.10107421875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 37630 + }, + { + "epoch": 30.05670926517572, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 37631 + }, + { + "epoch": 30.057507987220447, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 37632 + }, + { + "epoch": 30.058306709265175, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 37633 + }, + { + "epoch": 30.059105431309906, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 37634 + }, + { + "epoch": 30.059904153354633, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 37635 + }, + { + "epoch": 30.06070287539936, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37636 + }, + { + "epoch": 30.06150159744409, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 37637 + }, + { + "epoch": 30.062300319488816, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 37638 + }, + { + "epoch": 30.063099041533548, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 37639 + }, + { + "epoch": 30.063897763578275, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 37640 + }, + { + "epoch": 30.064696485623003, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 37641 + }, + { + "epoch": 30.06549520766773, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 37642 + }, + { + "epoch": 30.06629392971246, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 37643 + }, + { + "epoch": 30.06709265175719, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37644 + }, + { + "epoch": 30.067891373801917, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 37645 + }, + { + "epoch": 30.068690095846645, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37646 + }, + { + "epoch": 30.069488817891372, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 37647 + }, + { + "epoch": 30.070287539936103, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 37648 + }, + { + "epoch": 30.07108626198083, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 37649 + }, + { + "epoch": 30.07188498402556, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 37650 + }, + { + "epoch": 30.072683706070286, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 37651 + }, + { + "epoch": 30.073482428115017, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 37652 + }, + { + "epoch": 30.074281150159745, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 37653 + }, + { + "epoch": 30.075079872204473, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 37654 + }, + { + "epoch": 30.0758785942492, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 37655 + }, + { + "epoch": 30.076677316293928, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 37656 + }, + { + "epoch": 30.07747603833866, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 37657 + }, + { + "epoch": 30.078274760383387, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 37658 + }, + { + "epoch": 30.079073482428115, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 37659 + }, + { + "epoch": 30.079872204472842, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 37660 + }, + { + "epoch": 30.080670926517573, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 37661 + }, + { + "epoch": 30.0814696485623, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 37662 + }, + { + "epoch": 30.08226837060703, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.075, + "step": 37663 + }, + { + "epoch": 30.083067092651756, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37664 + }, + { + "epoch": 30.083865814696484, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 37665 + }, + { + "epoch": 30.084664536741215, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37666 + }, + { + "epoch": 30.085463258785943, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 37667 + }, + { + "epoch": 30.08626198083067, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 37668 + }, + { + "epoch": 30.087060702875398, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37669 + }, + { + "epoch": 30.08785942492013, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 37670 + }, + { + "epoch": 30.088658146964857, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37671 + }, + { + "epoch": 30.089456869009584, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 37672 + }, + { + "epoch": 30.090255591054312, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 37673 + }, + { + "epoch": 30.09105431309904, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 37674 + }, + { + "epoch": 30.09185303514377, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 37675 + }, + { + "epoch": 30.0926517571885, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 37676 + }, + { + "epoch": 30.093450479233226, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 37677 + }, + { + "epoch": 30.094249201277954, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 37678 + }, + { + "epoch": 30.095047923322685, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 37679 + }, + { + "epoch": 30.095846645367413, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.0761, + "step": 37680 + }, + { + "epoch": 30.09664536741214, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 37681 + }, + { + "epoch": 30.097444089456868, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37682 + }, + { + "epoch": 30.0982428115016, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 37683 + }, + { + "epoch": 30.099041533546327, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 37684 + }, + { + "epoch": 30.099840255591054, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 37685 + }, + { + "epoch": 30.100638977635782, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 37686 + }, + { + "epoch": 30.10143769968051, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 37687 + }, + { + "epoch": 30.10223642172524, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 37688 + }, + { + "epoch": 30.10303514376997, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 37689 + }, + { + "epoch": 30.103833865814696, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 37690 + }, + { + "epoch": 30.104632587859424, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 37691 + }, + { + "epoch": 30.105431309904155, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 37692 + }, + { + "epoch": 30.106230031948883, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 37693 + }, + { + "epoch": 30.10702875399361, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37694 + }, + { + "epoch": 30.107827476038338, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 37695 + }, + { + "epoch": 30.108626198083066, + "grad_norm": 0.10400390625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37696 + }, + { + "epoch": 30.109424920127797, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 37697 + }, + { + "epoch": 30.110223642172524, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.1029, + "step": 37698 + }, + { + "epoch": 30.111022364217252, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 37699 + }, + { + "epoch": 30.11182108626198, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 37700 + }, + { + "epoch": 30.11261980830671, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 37701 + }, + { + "epoch": 30.11341853035144, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 37702 + }, + { + "epoch": 30.114217252396166, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 37703 + }, + { + "epoch": 30.115015974440894, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 37704 + }, + { + "epoch": 30.11581469648562, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 37705 + }, + { + "epoch": 30.116613418530353, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 37706 + }, + { + "epoch": 30.11741214057508, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37707 + }, + { + "epoch": 30.118210862619808, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 37708 + }, + { + "epoch": 30.119009584664536, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37709 + }, + { + "epoch": 30.119808306709267, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 37710 + }, + { + "epoch": 30.120607028753994, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 37711 + }, + { + "epoch": 30.121405750798722, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 37712 + }, + { + "epoch": 30.12220447284345, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 37713 + }, + { + "epoch": 30.123003194888177, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37714 + }, + { + "epoch": 30.12380191693291, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 37715 + }, + { + "epoch": 30.124600638977636, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 37716 + }, + { + "epoch": 30.125399361022364, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 37717 + }, + { + "epoch": 30.12619808306709, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 37718 + }, + { + "epoch": 30.126996805111823, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 37719 + }, + { + "epoch": 30.12779552715655, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0742, + "step": 37720 + }, + { + "epoch": 30.128594249201278, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 37721 + }, + { + "epoch": 30.129392971246006, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 37722 + }, + { + "epoch": 30.130191693290733, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 37723 + }, + { + "epoch": 30.130990415335464, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 37724 + }, + { + "epoch": 30.131789137380192, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 37725 + }, + { + "epoch": 30.13258785942492, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37726 + }, + { + "epoch": 30.133386581469647, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37727 + }, + { + "epoch": 30.13418530351438, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 37728 + }, + { + "epoch": 30.134984025559106, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 37729 + }, + { + "epoch": 30.135782747603834, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 37730 + }, + { + "epoch": 30.13658146964856, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37731 + }, + { + "epoch": 30.13738019169329, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 37732 + }, + { + "epoch": 30.13817891373802, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 37733 + }, + { + "epoch": 30.138977635782748, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 37734 + }, + { + "epoch": 30.139776357827476, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 37735 + }, + { + "epoch": 30.140575079872203, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 37736 + }, + { + "epoch": 30.141373801916934, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 37737 + }, + { + "epoch": 30.142172523961662, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 37738 + }, + { + "epoch": 30.14297124600639, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 37739 + }, + { + "epoch": 30.143769968051117, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 37740 + }, + { + "epoch": 30.144568690095845, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 37741 + }, + { + "epoch": 30.145367412140576, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 37742 + }, + { + "epoch": 30.146166134185304, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 37743 + }, + { + "epoch": 30.14696485623003, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 37744 + }, + { + "epoch": 30.14776357827476, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 37745 + }, + { + "epoch": 30.14856230031949, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 37746 + }, + { + "epoch": 30.149361022364218, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 37747 + }, + { + "epoch": 30.150159744408946, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 37748 + }, + { + "epoch": 30.150958466453673, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 37749 + }, + { + "epoch": 30.1517571884984, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37750 + }, + { + "epoch": 30.152555910543132, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 37751 + }, + { + "epoch": 30.15335463258786, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 37752 + }, + { + "epoch": 30.154153354632587, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 37753 + }, + { + "epoch": 30.154952076677315, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 37754 + }, + { + "epoch": 30.155750798722046, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 37755 + }, + { + "epoch": 30.156549520766774, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 37756 + }, + { + "epoch": 30.1573482428115, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 37757 + }, + { + "epoch": 30.15814696485623, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 37758 + }, + { + "epoch": 30.15894568690096, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37759 + }, + { + "epoch": 30.159744408945688, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 37760 + }, + { + "epoch": 30.160543130990416, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 37761 + }, + { + "epoch": 30.161341853035143, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 37762 + }, + { + "epoch": 30.16214057507987, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 37763 + }, + { + "epoch": 30.162939297124602, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 37764 + }, + { + "epoch": 30.16373801916933, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37765 + }, + { + "epoch": 30.164536741214057, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 37766 + }, + { + "epoch": 30.165335463258785, + "grad_norm": 0.09716796875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 37767 + }, + { + "epoch": 30.166134185303516, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37768 + }, + { + "epoch": 30.166932907348244, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 37769 + }, + { + "epoch": 30.16773162939297, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37770 + }, + { + "epoch": 30.1685303514377, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 37771 + }, + { + "epoch": 30.169329073482427, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 37772 + }, + { + "epoch": 30.170127795527158, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 37773 + }, + { + "epoch": 30.170926517571885, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37774 + }, + { + "epoch": 30.171725239616613, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 37775 + }, + { + "epoch": 30.17252396166134, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 37776 + }, + { + "epoch": 30.173322683706072, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37777 + }, + { + "epoch": 30.1741214057508, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37778 + }, + { + "epoch": 30.174920127795527, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37779 + }, + { + "epoch": 30.175718849840255, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 37780 + }, + { + "epoch": 30.176517571884983, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37781 + }, + { + "epoch": 30.177316293929714, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 37782 + }, + { + "epoch": 30.17811501597444, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0943, + "step": 37783 + }, + { + "epoch": 30.17891373801917, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 37784 + }, + { + "epoch": 30.179712460063897, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 37785 + }, + { + "epoch": 30.180511182108628, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 37786 + }, + { + "epoch": 30.181309904153355, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 37787 + }, + { + "epoch": 30.182108626198083, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 37788 + }, + { + "epoch": 30.18290734824281, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 37789 + }, + { + "epoch": 30.18370607028754, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37790 + }, + { + "epoch": 30.18450479233227, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 37791 + }, + { + "epoch": 30.185303514376997, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 37792 + }, + { + "epoch": 30.186102236421725, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 37793 + }, + { + "epoch": 30.186900958466452, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 37794 + }, + { + "epoch": 30.187699680511184, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 37795 + }, + { + "epoch": 30.18849840255591, + "grad_norm": 0.09521484375, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 37796 + }, + { + "epoch": 30.18929712460064, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 37797 + }, + { + "epoch": 30.190095846645367, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37798 + }, + { + "epoch": 30.190894568690094, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 37799 + }, + { + "epoch": 30.191693290734825, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0756, + "step": 37800 + }, + { + "epoch": 30.192492012779553, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 37801 + }, + { + "epoch": 30.19329073482428, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 37802 + }, + { + "epoch": 30.19408945686901, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 37803 + }, + { + "epoch": 30.19488817891374, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 37804 + }, + { + "epoch": 30.195686900958467, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 37805 + }, + { + "epoch": 30.196485623003195, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 37806 + }, + { + "epoch": 30.197284345047922, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 37807 + }, + { + "epoch": 30.19808306709265, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 37808 + }, + { + "epoch": 30.19888178913738, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 37809 + }, + { + "epoch": 30.19968051118211, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 37810 + }, + { + "epoch": 30.200479233226837, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 37811 + }, + { + "epoch": 30.201277955271564, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 37812 + }, + { + "epoch": 30.202076677316295, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 37813 + }, + { + "epoch": 30.202875399361023, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37814 + }, + { + "epoch": 30.20367412140575, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 37815 + }, + { + "epoch": 30.20447284345048, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 37816 + }, + { + "epoch": 30.205271565495206, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37817 + }, + { + "epoch": 30.206070287539937, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 37818 + }, + { + "epoch": 30.206869009584665, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 37819 + }, + { + "epoch": 30.207667731629392, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 37820 + }, + { + "epoch": 30.20846645367412, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 37821 + }, + { + "epoch": 30.20926517571885, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 37822 + }, + { + "epoch": 30.21006389776358, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 37823 + }, + { + "epoch": 30.210862619808307, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 37824 + }, + { + "epoch": 30.211661341853034, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 37825 + }, + { + "epoch": 30.212460063897762, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37826 + }, + { + "epoch": 30.213258785942493, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 37827 + }, + { + "epoch": 30.21405750798722, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 37828 + }, + { + "epoch": 30.21485623003195, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 37829 + }, + { + "epoch": 30.215654952076676, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 37830 + }, + { + "epoch": 30.216453674121407, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 37831 + }, + { + "epoch": 30.217252396166135, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 37832 + }, + { + "epoch": 30.218051118210862, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37833 + }, + { + "epoch": 30.21884984025559, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 37834 + }, + { + "epoch": 30.21964856230032, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 37835 + }, + { + "epoch": 30.22044728434505, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 37836 + }, + { + "epoch": 30.221246006389777, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 37837 + }, + { + "epoch": 30.222044728434504, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 37838 + }, + { + "epoch": 30.222843450479232, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 37839 + }, + { + "epoch": 30.223642172523963, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 37840 + }, + { + "epoch": 30.22444089456869, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 37841 + }, + { + "epoch": 30.22523961661342, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 37842 + }, + { + "epoch": 30.226038338658146, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 37843 + }, + { + "epoch": 30.226837060702877, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 37844 + }, + { + "epoch": 30.227635782747605, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 37845 + }, + { + "epoch": 30.228434504792332, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 37846 + }, + { + "epoch": 30.22923322683706, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 37847 + }, + { + "epoch": 30.230031948881788, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 37848 + }, + { + "epoch": 30.23083067092652, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 37849 + }, + { + "epoch": 30.231629392971247, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 37850 + }, + { + "epoch": 30.232428115015974, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 37851 + }, + { + "epoch": 30.233226837060702, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 37852 + }, + { + "epoch": 30.234025559105433, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 37853 + }, + { + "epoch": 30.23482428115016, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 37854 + }, + { + "epoch": 30.23562300319489, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 37855 + }, + { + "epoch": 30.236421725239616, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 37856 + }, + { + "epoch": 30.237220447284344, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 37857 + }, + { + "epoch": 30.238019169329075, + "grad_norm": 0.1005859375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 37858 + }, + { + "epoch": 30.238817891373802, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 37859 + }, + { + "epoch": 30.23961661341853, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 37860 + }, + { + "epoch": 30.240415335463258, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 37861 + }, + { + "epoch": 30.24121405750799, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37862 + }, + { + "epoch": 30.242012779552716, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 37863 + }, + { + "epoch": 30.242811501597444, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 37864 + }, + { + "epoch": 30.24361022364217, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 37865 + }, + { + "epoch": 30.2444089456869, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 37866 + }, + { + "epoch": 30.24520766773163, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 37867 + }, + { + "epoch": 30.24600638977636, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0978, + "step": 37868 + }, + { + "epoch": 30.246805111821086, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 37869 + }, + { + "epoch": 30.247603833865814, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 37870 + }, + { + "epoch": 30.248402555910545, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 37871 + }, + { + "epoch": 30.249201277955272, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 37872 + }, + { + "epoch": 30.25, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 37873 + }, + { + "epoch": 30.250798722044728, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 37874 + }, + { + "epoch": 30.251597444089455, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 37875 + }, + { + "epoch": 30.252396166134186, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 37876 + }, + { + "epoch": 30.253194888178914, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0749, + "step": 37877 + }, + { + "epoch": 30.25399361022364, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 37878 + }, + { + "epoch": 30.25479233226837, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 37879 + }, + { + "epoch": 30.2555910543131, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 37880 + }, + { + "epoch": 30.25638977635783, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 37881 + }, + { + "epoch": 30.257188498402556, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 37882 + }, + { + "epoch": 30.257987220447284, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 37883 + }, + { + "epoch": 30.25878594249201, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 37884 + }, + { + "epoch": 30.259584664536742, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 37885 + }, + { + "epoch": 30.26038338658147, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 37886 + }, + { + "epoch": 30.261182108626198, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 37887 + }, + { + "epoch": 30.261980830670925, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 37888 + }, + { + "epoch": 30.262779552715656, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 37889 + }, + { + "epoch": 30.263578274760384, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37890 + }, + { + "epoch": 30.26437699680511, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 37891 + }, + { + "epoch": 30.26517571884984, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 37892 + }, + { + "epoch": 30.265974440894567, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 37893 + }, + { + "epoch": 30.266773162939298, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 37894 + }, + { + "epoch": 30.267571884984026, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 37895 + }, + { + "epoch": 30.268370607028753, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 37896 + }, + { + "epoch": 30.26916932907348, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 37897 + }, + { + "epoch": 30.269968051118212, + "grad_norm": 0.240234375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 37898 + }, + { + "epoch": 30.27076677316294, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 37899 + }, + { + "epoch": 30.271565495207668, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 37900 + }, + { + "epoch": 30.272364217252395, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 37901 + }, + { + "epoch": 30.273162939297123, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 37902 + }, + { + "epoch": 30.273961661341854, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 37903 + }, + { + "epoch": 30.27476038338658, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 37904 + }, + { + "epoch": 30.27555910543131, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 37905 + }, + { + "epoch": 30.276357827476037, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.1008, + "step": 37906 + }, + { + "epoch": 30.277156549520768, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 37907 + }, + { + "epoch": 30.277955271565496, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 37908 + }, + { + "epoch": 30.278753993610223, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 37909 + }, + { + "epoch": 30.27955271565495, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 37910 + }, + { + "epoch": 30.28035143769968, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 37911 + }, + { + "epoch": 30.28115015974441, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 37912 + }, + { + "epoch": 30.281948881789138, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 37913 + }, + { + "epoch": 30.282747603833865, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 37914 + }, + { + "epoch": 30.283546325878593, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 37915 + }, + { + "epoch": 30.284345047923324, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 37916 + }, + { + "epoch": 30.28514376996805, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0761, + "step": 37917 + }, + { + "epoch": 30.28594249201278, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 37918 + }, + { + "epoch": 30.286741214057507, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 37919 + }, + { + "epoch": 30.287539936102238, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37920 + }, + { + "epoch": 30.288338658146966, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 37921 + }, + { + "epoch": 30.289137380191693, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 37922 + }, + { + "epoch": 30.28993610223642, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 37923 + }, + { + "epoch": 30.29073482428115, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 37924 + }, + { + "epoch": 30.29153354632588, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 37925 + }, + { + "epoch": 30.292332268370608, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 37926 + }, + { + "epoch": 30.293130990415335, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 37927 + }, + { + "epoch": 30.293929712460063, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 37928 + }, + { + "epoch": 30.294728434504794, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 37929 + }, + { + "epoch": 30.29552715654952, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37930 + }, + { + "epoch": 30.29632587859425, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 37931 + }, + { + "epoch": 30.297124600638977, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37932 + }, + { + "epoch": 30.297923322683705, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 37933 + }, + { + "epoch": 30.298722044728436, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 37934 + }, + { + "epoch": 30.299520766773163, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 37935 + }, + { + "epoch": 30.30031948881789, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 37936 + }, + { + "epoch": 30.30111821086262, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 37937 + }, + { + "epoch": 30.30191693290735, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 37938 + }, + { + "epoch": 30.302715654952078, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 37939 + }, + { + "epoch": 30.303514376996805, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 37940 + }, + { + "epoch": 30.304313099041533, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 37941 + }, + { + "epoch": 30.30511182108626, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 37942 + }, + { + "epoch": 30.30591054313099, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 37943 + }, + { + "epoch": 30.30670926517572, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 37944 + }, + { + "epoch": 30.307507987220447, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 37945 + }, + { + "epoch": 30.308306709265175, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 37946 + }, + { + "epoch": 30.309105431309906, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 37947 + }, + { + "epoch": 30.309904153354633, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 37948 + }, + { + "epoch": 30.31070287539936, + "grad_norm": 0.26171875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 37949 + }, + { + "epoch": 30.31150159744409, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 37950 + }, + { + "epoch": 30.312300319488816, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 37951 + }, + { + "epoch": 30.313099041533548, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 37952 + }, + { + "epoch": 30.313897763578275, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 37953 + }, + { + "epoch": 30.314696485623003, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 37954 + }, + { + "epoch": 30.31549520766773, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 37955 + }, + { + "epoch": 30.31629392971246, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0933, + "step": 37956 + }, + { + "epoch": 30.31709265175719, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 37957 + }, + { + "epoch": 30.317891373801917, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 37958 + }, + { + "epoch": 30.318690095846645, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 37959 + }, + { + "epoch": 30.319488817891372, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 37960 + }, + { + "epoch": 30.320287539936103, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 37961 + }, + { + "epoch": 30.32108626198083, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 37962 + }, + { + "epoch": 30.32188498402556, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 37963 + }, + { + "epoch": 30.322683706070286, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 37964 + }, + { + "epoch": 30.323482428115017, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 37965 + }, + { + "epoch": 30.324281150159745, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 37966 + }, + { + "epoch": 30.325079872204473, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 37967 + }, + { + "epoch": 30.3258785942492, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 37968 + }, + { + "epoch": 30.326677316293928, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 37969 + }, + { + "epoch": 30.32747603833866, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 37970 + }, + { + "epoch": 30.328274760383387, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 37971 + }, + { + "epoch": 30.329073482428115, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 37972 + }, + { + "epoch": 30.329872204472842, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 37973 + }, + { + "epoch": 30.330670926517573, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 37974 + }, + { + "epoch": 30.3314696485623, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 37975 + }, + { + "epoch": 30.33226837060703, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 37976 + }, + { + "epoch": 30.333067092651756, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 37977 + }, + { + "epoch": 30.333865814696484, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 37978 + }, + { + "epoch": 30.334664536741215, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 37979 + }, + { + "epoch": 30.335463258785943, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 37980 + }, + { + "epoch": 30.33626198083067, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37981 + }, + { + "epoch": 30.337060702875398, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 37982 + }, + { + "epoch": 30.33785942492013, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 37983 + }, + { + "epoch": 30.338658146964857, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 37984 + }, + { + "epoch": 30.339456869009584, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 37985 + }, + { + "epoch": 30.340255591054312, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 37986 + }, + { + "epoch": 30.341054313099043, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 37987 + }, + { + "epoch": 30.34185303514377, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 37988 + }, + { + "epoch": 30.3426517571885, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 37989 + }, + { + "epoch": 30.343450479233226, + "grad_norm": 0.099609375, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 37990 + }, + { + "epoch": 30.344249201277954, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 37991 + }, + { + "epoch": 30.345047923322685, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37992 + }, + { + "epoch": 30.345846645367413, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 37993 + }, + { + "epoch": 30.34664536741214, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 37994 + }, + { + "epoch": 30.347444089456868, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 37995 + }, + { + "epoch": 30.3482428115016, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 37996 + }, + { + "epoch": 30.349041533546327, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 37997 + }, + { + "epoch": 30.349840255591054, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 37998 + }, + { + "epoch": 30.350638977635782, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 37999 + }, + { + "epoch": 30.35143769968051, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 38000 + }, + { + "epoch": 30.35223642172524, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 38001 + }, + { + "epoch": 30.35303514376997, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 38002 + }, + { + "epoch": 30.353833865814696, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38003 + }, + { + "epoch": 30.354632587859424, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0722, + "step": 38004 + }, + { + "epoch": 30.355431309904155, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 38005 + }, + { + "epoch": 30.356230031948883, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 38006 + }, + { + "epoch": 30.35702875399361, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 38007 + }, + { + "epoch": 30.357827476038338, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 38008 + }, + { + "epoch": 30.358626198083066, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 38009 + }, + { + "epoch": 30.359424920127797, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.074, + "step": 38010 + }, + { + "epoch": 30.360223642172524, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 38011 + }, + { + "epoch": 30.361022364217252, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38012 + }, + { + "epoch": 30.36182108626198, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 38013 + }, + { + "epoch": 30.36261980830671, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 38014 + }, + { + "epoch": 30.36341853035144, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 38015 + }, + { + "epoch": 30.364217252396166, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 38016 + }, + { + "epoch": 30.365015974440894, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38017 + }, + { + "epoch": 30.36581469648562, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 38018 + }, + { + "epoch": 30.366613418530353, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 38019 + }, + { + "epoch": 30.36741214057508, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 38020 + }, + { + "epoch": 30.368210862619808, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 38021 + }, + { + "epoch": 30.369009584664536, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 38022 + }, + { + "epoch": 30.369808306709267, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 38023 + }, + { + "epoch": 30.370607028753994, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 38024 + }, + { + "epoch": 30.371405750798722, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38025 + }, + { + "epoch": 30.37220447284345, + "grad_norm": 0.10546875, + "learning_rate": 0.0001, + "loss": 1.0963, + "step": 38026 + }, + { + "epoch": 30.373003194888177, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 38027 + }, + { + "epoch": 30.37380191693291, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38028 + }, + { + "epoch": 30.374600638977636, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 38029 + }, + { + "epoch": 30.375399361022364, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 38030 + }, + { + "epoch": 30.37619808306709, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 38031 + }, + { + "epoch": 30.376996805111823, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 38032 + }, + { + "epoch": 30.37779552715655, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 38033 + }, + { + "epoch": 30.378594249201278, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 38034 + }, + { + "epoch": 30.379392971246006, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 38035 + }, + { + "epoch": 30.380191693290733, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38036 + }, + { + "epoch": 30.380990415335464, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 38037 + }, + { + "epoch": 30.381789137380192, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 38038 + }, + { + "epoch": 30.38258785942492, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 38039 + }, + { + "epoch": 30.383386581469647, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 38040 + }, + { + "epoch": 30.38418530351438, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 38041 + }, + { + "epoch": 30.384984025559106, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 38042 + }, + { + "epoch": 30.385782747603834, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 38043 + }, + { + "epoch": 30.38658146964856, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 38044 + }, + { + "epoch": 30.38738019169329, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 38045 + }, + { + "epoch": 30.38817891373802, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 38046 + }, + { + "epoch": 30.388977635782748, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 38047 + }, + { + "epoch": 30.389776357827476, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38048 + }, + { + "epoch": 30.390575079872203, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 38049 + }, + { + "epoch": 30.391373801916934, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 38050 + }, + { + "epoch": 30.392172523961662, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38051 + }, + { + "epoch": 30.39297124600639, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 38052 + }, + { + "epoch": 30.393769968051117, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 38053 + }, + { + "epoch": 30.394568690095845, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 38054 + }, + { + "epoch": 30.395367412140576, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 38055 + }, + { + "epoch": 30.396166134185304, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 38056 + }, + { + "epoch": 30.39696485623003, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 38057 + }, + { + "epoch": 30.39776357827476, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 38058 + }, + { + "epoch": 30.39856230031949, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 38059 + }, + { + "epoch": 30.399361022364218, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 38060 + }, + { + "epoch": 30.400159744408946, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 38061 + }, + { + "epoch": 30.400958466453673, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 38062 + }, + { + "epoch": 30.4017571884984, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 38063 + }, + { + "epoch": 30.402555910543132, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 38064 + }, + { + "epoch": 30.40335463258786, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 38065 + }, + { + "epoch": 30.404153354632587, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 38066 + }, + { + "epoch": 30.404952076677315, + "grad_norm": 0.109375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 38067 + }, + { + "epoch": 30.405750798722046, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 38068 + }, + { + "epoch": 30.406549520766774, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 38069 + }, + { + "epoch": 30.4073482428115, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 38070 + }, + { + "epoch": 30.40814696485623, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 38071 + }, + { + "epoch": 30.408945686900957, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 38072 + }, + { + "epoch": 30.409744408945688, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 38073 + }, + { + "epoch": 30.410543130990416, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 38074 + }, + { + "epoch": 30.411341853035143, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 38075 + }, + { + "epoch": 30.41214057507987, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 38076 + }, + { + "epoch": 30.412939297124602, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38077 + }, + { + "epoch": 30.41373801916933, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 38078 + }, + { + "epoch": 30.414536741214057, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 38079 + }, + { + "epoch": 30.415335463258785, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38080 + }, + { + "epoch": 30.416134185303516, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 38081 + }, + { + "epoch": 30.416932907348244, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38082 + }, + { + "epoch": 30.41773162939297, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 38083 + }, + { + "epoch": 30.4185303514377, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38084 + }, + { + "epoch": 30.419329073482427, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 38085 + }, + { + "epoch": 30.420127795527158, + "grad_norm": 0.251953125, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 38086 + }, + { + "epoch": 30.420926517571885, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 38087 + }, + { + "epoch": 30.421725239616613, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 38088 + }, + { + "epoch": 30.42252396166134, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 38089 + }, + { + "epoch": 30.423322683706072, + "grad_norm": 0.2392578125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 38090 + }, + { + "epoch": 30.4241214057508, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 38091 + }, + { + "epoch": 30.424920127795527, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 38092 + }, + { + "epoch": 30.425718849840255, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 38093 + }, + { + "epoch": 30.426517571884983, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 38094 + }, + { + "epoch": 30.427316293929714, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 38095 + }, + { + "epoch": 30.42811501597444, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 38096 + }, + { + "epoch": 30.42891373801917, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 38097 + }, + { + "epoch": 30.429712460063897, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 38098 + }, + { + "epoch": 30.430511182108628, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 38099 + }, + { + "epoch": 30.431309904153355, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 38100 + }, + { + "epoch": 30.432108626198083, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 38101 + }, + { + "epoch": 30.43290734824281, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0746, + "step": 38102 + }, + { + "epoch": 30.43370607028754, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 38103 + }, + { + "epoch": 30.43450479233227, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 38104 + }, + { + "epoch": 30.435303514376997, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 38105 + }, + { + "epoch": 30.436102236421725, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 38106 + }, + { + "epoch": 30.436900958466452, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 38107 + }, + { + "epoch": 30.437699680511184, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 38108 + }, + { + "epoch": 30.43849840255591, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 38109 + }, + { + "epoch": 30.43929712460064, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 38110 + }, + { + "epoch": 30.440095846645367, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 38111 + }, + { + "epoch": 30.440894568690094, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 38112 + }, + { + "epoch": 30.441693290734825, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 38113 + }, + { + "epoch": 30.442492012779553, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 38114 + }, + { + "epoch": 30.44329073482428, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 38115 + }, + { + "epoch": 30.44408945686901, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 38116 + }, + { + "epoch": 30.44488817891374, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1036, + "step": 38117 + }, + { + "epoch": 30.445686900958467, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 38118 + }, + { + "epoch": 30.446485623003195, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 38119 + }, + { + "epoch": 30.447284345047922, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 38120 + }, + { + "epoch": 30.44808306709265, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 38121 + }, + { + "epoch": 30.44888178913738, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0985, + "step": 38122 + }, + { + "epoch": 30.44968051118211, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 38123 + }, + { + "epoch": 30.450479233226837, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 38124 + }, + { + "epoch": 30.451277955271564, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 38125 + }, + { + "epoch": 30.452076677316295, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 38126 + }, + { + "epoch": 30.452875399361023, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38127 + }, + { + "epoch": 30.45367412140575, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 38128 + }, + { + "epoch": 30.45447284345048, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 38129 + }, + { + "epoch": 30.455271565495206, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 38130 + }, + { + "epoch": 30.456070287539937, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 38131 + }, + { + "epoch": 30.456869009584665, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38132 + }, + { + "epoch": 30.457667731629392, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 38133 + }, + { + "epoch": 30.45846645367412, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 38134 + }, + { + "epoch": 30.45926517571885, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 38135 + }, + { + "epoch": 30.46006389776358, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 38136 + }, + { + "epoch": 30.460862619808307, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 38137 + }, + { + "epoch": 30.461661341853034, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 38138 + }, + { + "epoch": 30.462460063897762, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0751, + "step": 38139 + }, + { + "epoch": 30.463258785942493, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 38140 + }, + { + "epoch": 30.46405750798722, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 38141 + }, + { + "epoch": 30.46485623003195, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 38142 + }, + { + "epoch": 30.465654952076676, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 38143 + }, + { + "epoch": 30.466453674121407, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 38144 + }, + { + "epoch": 30.467252396166135, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 38145 + }, + { + "epoch": 30.468051118210862, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 38146 + }, + { + "epoch": 30.46884984025559, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 38147 + }, + { + "epoch": 30.46964856230032, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 38148 + }, + { + "epoch": 30.47044728434505, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 38149 + }, + { + "epoch": 30.471246006389777, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 38150 + }, + { + "epoch": 30.472044728434504, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 38151 + }, + { + "epoch": 30.472843450479232, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 38152 + }, + { + "epoch": 30.473642172523963, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 38153 + }, + { + "epoch": 30.47444089456869, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 38154 + }, + { + "epoch": 30.47523961661342, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 38155 + }, + { + "epoch": 30.476038338658146, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 38156 + }, + { + "epoch": 30.476837060702877, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 38157 + }, + { + "epoch": 30.477635782747605, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 38158 + }, + { + "epoch": 30.478434504792332, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 38159 + }, + { + "epoch": 30.47923322683706, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 38160 + }, + { + "epoch": 30.480031948881788, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 38161 + }, + { + "epoch": 30.48083067092652, + "grad_norm": 0.1044921875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 38162 + }, + { + "epoch": 30.481629392971247, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 38163 + }, + { + "epoch": 30.482428115015974, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 38164 + }, + { + "epoch": 30.483226837060702, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38165 + }, + { + "epoch": 30.484025559105433, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 38166 + }, + { + "epoch": 30.48482428115016, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 38167 + }, + { + "epoch": 30.48562300319489, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 38168 + }, + { + "epoch": 30.486421725239616, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 38169 + }, + { + "epoch": 30.487220447284344, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 38170 + }, + { + "epoch": 30.488019169329075, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 38171 + }, + { + "epoch": 30.488817891373802, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 38172 + }, + { + "epoch": 30.48961661341853, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 38173 + }, + { + "epoch": 30.490415335463258, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 38174 + }, + { + "epoch": 30.49121405750799, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 38175 + }, + { + "epoch": 30.492012779552716, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 38176 + }, + { + "epoch": 30.492811501597444, + "grad_norm": 0.2216796875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 38177 + }, + { + "epoch": 30.49361022364217, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 38178 + }, + { + "epoch": 30.4944089456869, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 38179 + }, + { + "epoch": 30.49520766773163, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 38180 + }, + { + "epoch": 30.49600638977636, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 38181 + }, + { + "epoch": 30.496805111821086, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0763, + "step": 38182 + }, + { + "epoch": 30.497603833865814, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38183 + }, + { + "epoch": 30.498402555910545, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 38184 + }, + { + "epoch": 30.499201277955272, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 38185 + }, + { + "epoch": 30.5, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 38186 + }, + { + "epoch": 30.500798722044728, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 38187 + }, + { + "epoch": 30.501597444089455, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 38188 + }, + { + "epoch": 30.502396166134186, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 38189 + }, + { + "epoch": 30.503194888178914, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 38190 + }, + { + "epoch": 30.50399361022364, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 38191 + }, + { + "epoch": 30.50479233226837, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38192 + }, + { + "epoch": 30.5055910543131, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 38193 + }, + { + "epoch": 30.50638977635783, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 38194 + }, + { + "epoch": 30.507188498402556, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 38195 + }, + { + "epoch": 30.507987220447284, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 38196 + }, + { + "epoch": 30.50878594249201, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 38197 + }, + { + "epoch": 30.509584664536742, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 38198 + }, + { + "epoch": 30.51038338658147, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38199 + }, + { + "epoch": 30.511182108626198, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 38200 + }, + { + "epoch": 30.511980830670925, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 38201 + }, + { + "epoch": 30.512779552715656, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 38202 + }, + { + "epoch": 30.513578274760384, + "grad_norm": 0.263671875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38203 + }, + { + "epoch": 30.51437699680511, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38204 + }, + { + "epoch": 30.51517571884984, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 38205 + }, + { + "epoch": 30.515974440894567, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 38206 + }, + { + "epoch": 30.516773162939298, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 38207 + }, + { + "epoch": 30.517571884984026, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38208 + }, + { + "epoch": 30.518370607028753, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 38209 + }, + { + "epoch": 30.51916932907348, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 38210 + }, + { + "epoch": 30.519968051118212, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 38211 + }, + { + "epoch": 30.52076677316294, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 38212 + }, + { + "epoch": 30.521565495207668, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 38213 + }, + { + "epoch": 30.522364217252395, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 38214 + }, + { + "epoch": 30.523162939297123, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 38215 + }, + { + "epoch": 30.523961661341854, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 38216 + }, + { + "epoch": 30.52476038338658, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 38217 + }, + { + "epoch": 30.52555910543131, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1001, + "step": 38218 + }, + { + "epoch": 30.526357827476037, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38219 + }, + { + "epoch": 30.527156549520768, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 38220 + }, + { + "epoch": 30.527955271565496, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 38221 + }, + { + "epoch": 30.528753993610223, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 38222 + }, + { + "epoch": 30.52955271565495, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 38223 + }, + { + "epoch": 30.53035143769968, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 38224 + }, + { + "epoch": 30.53115015974441, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 38225 + }, + { + "epoch": 30.531948881789138, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 38226 + }, + { + "epoch": 30.532747603833865, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 38227 + }, + { + "epoch": 30.533546325878593, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38228 + }, + { + "epoch": 30.534345047923324, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 38229 + }, + { + "epoch": 30.53514376996805, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 38230 + }, + { + "epoch": 30.53594249201278, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 38231 + }, + { + "epoch": 30.536741214057507, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 38232 + }, + { + "epoch": 30.537539936102235, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 38233 + }, + { + "epoch": 30.538338658146966, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 38234 + }, + { + "epoch": 30.539137380191693, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0741, + "step": 38235 + }, + { + "epoch": 30.53993610223642, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 38236 + }, + { + "epoch": 30.54073482428115, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 38237 + }, + { + "epoch": 30.54153354632588, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 38238 + }, + { + "epoch": 30.542332268370608, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 38239 + }, + { + "epoch": 30.543130990415335, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 38240 + }, + { + "epoch": 30.543929712460063, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 38241 + }, + { + "epoch": 30.544728434504794, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 38242 + }, + { + "epoch": 30.54552715654952, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 38243 + }, + { + "epoch": 30.54632587859425, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 38244 + }, + { + "epoch": 30.547124600638977, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 38245 + }, + { + "epoch": 30.547923322683705, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 38246 + }, + { + "epoch": 30.548722044728436, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 38247 + }, + { + "epoch": 30.549520766773163, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 38248 + }, + { + "epoch": 30.55031948881789, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38249 + }, + { + "epoch": 30.55111821086262, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0968, + "step": 38250 + }, + { + "epoch": 30.55191693290735, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38251 + }, + { + "epoch": 30.552715654952078, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 38252 + }, + { + "epoch": 30.553514376996805, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 38253 + }, + { + "epoch": 30.554313099041533, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 38254 + }, + { + "epoch": 30.55511182108626, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 38255 + }, + { + "epoch": 30.55591054313099, + "grad_norm": 0.2294921875, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 38256 + }, + { + "epoch": 30.55670926517572, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 38257 + }, + { + "epoch": 30.557507987220447, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 38258 + }, + { + "epoch": 30.558306709265175, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38259 + }, + { + "epoch": 30.559105431309906, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38260 + }, + { + "epoch": 30.559904153354633, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 38261 + }, + { + "epoch": 30.56070287539936, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 38262 + }, + { + "epoch": 30.56150159744409, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 38263 + }, + { + "epoch": 30.562300319488816, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 38264 + }, + { + "epoch": 30.563099041533548, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 38265 + }, + { + "epoch": 30.563897763578275, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38266 + }, + { + "epoch": 30.564696485623003, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 38267 + }, + { + "epoch": 30.56549520766773, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 38268 + }, + { + "epoch": 30.56629392971246, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 38269 + }, + { + "epoch": 30.56709265175719, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 38270 + }, + { + "epoch": 30.567891373801917, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0761, + "step": 38271 + }, + { + "epoch": 30.568690095846645, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 38272 + }, + { + "epoch": 30.569488817891372, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 38273 + }, + { + "epoch": 30.570287539936103, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 38274 + }, + { + "epoch": 30.57108626198083, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 38275 + }, + { + "epoch": 30.57188498402556, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 38276 + }, + { + "epoch": 30.572683706070286, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 38277 + }, + { + "epoch": 30.573482428115017, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 38278 + }, + { + "epoch": 30.574281150159745, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 38279 + }, + { + "epoch": 30.575079872204473, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 38280 + }, + { + "epoch": 30.5758785942492, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38281 + }, + { + "epoch": 30.576677316293928, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 38282 + }, + { + "epoch": 30.57747603833866, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 38283 + }, + { + "epoch": 30.578274760383387, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 38284 + }, + { + "epoch": 30.579073482428115, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38285 + }, + { + "epoch": 30.579872204472842, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 38286 + }, + { + "epoch": 30.580670926517573, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 38287 + }, + { + "epoch": 30.5814696485623, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0765, + "step": 38288 + }, + { + "epoch": 30.58226837060703, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38289 + }, + { + "epoch": 30.583067092651756, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 38290 + }, + { + "epoch": 30.583865814696484, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 38291 + }, + { + "epoch": 30.584664536741215, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 38292 + }, + { + "epoch": 30.585463258785943, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 38293 + }, + { + "epoch": 30.58626198083067, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 38294 + }, + { + "epoch": 30.587060702875398, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38295 + }, + { + "epoch": 30.58785942492013, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 38296 + }, + { + "epoch": 30.588658146964857, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38297 + }, + { + "epoch": 30.589456869009584, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0957, + "step": 38298 + }, + { + "epoch": 30.590255591054312, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 38299 + }, + { + "epoch": 30.591054313099043, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 38300 + }, + { + "epoch": 30.59185303514377, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 38301 + }, + { + "epoch": 30.5926517571885, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 38302 + }, + { + "epoch": 30.593450479233226, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 38303 + }, + { + "epoch": 30.594249201277954, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 38304 + }, + { + "epoch": 30.595047923322685, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 38305 + }, + { + "epoch": 30.595846645367413, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 38306 + }, + { + "epoch": 30.59664536741214, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 38307 + }, + { + "epoch": 30.597444089456868, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 38308 + }, + { + "epoch": 30.5982428115016, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 38309 + }, + { + "epoch": 30.599041533546327, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38310 + }, + { + "epoch": 30.599840255591054, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 38311 + }, + { + "epoch": 30.600638977635782, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 38312 + }, + { + "epoch": 30.60143769968051, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 38313 + }, + { + "epoch": 30.60223642172524, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 38314 + }, + { + "epoch": 30.60303514376997, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 38315 + }, + { + "epoch": 30.603833865814696, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 38316 + }, + { + "epoch": 30.604632587859424, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 38317 + }, + { + "epoch": 30.605431309904155, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 38318 + }, + { + "epoch": 30.606230031948883, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 38319 + }, + { + "epoch": 30.60702875399361, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0758, + "step": 38320 + }, + { + "epoch": 30.607827476038338, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 38321 + }, + { + "epoch": 30.608626198083066, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 38322 + }, + { + "epoch": 30.609424920127797, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 38323 + }, + { + "epoch": 30.610223642172524, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 38324 + }, + { + "epoch": 30.611022364217252, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 38325 + }, + { + "epoch": 30.61182108626198, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 38326 + }, + { + "epoch": 30.61261980830671, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 38327 + }, + { + "epoch": 30.61341853035144, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 38328 + }, + { + "epoch": 30.614217252396166, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 38329 + }, + { + "epoch": 30.615015974440894, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 38330 + }, + { + "epoch": 30.61581469648562, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38331 + }, + { + "epoch": 30.616613418530353, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 38332 + }, + { + "epoch": 30.61741214057508, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 38333 + }, + { + "epoch": 30.618210862619808, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 38334 + }, + { + "epoch": 30.619009584664536, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 38335 + }, + { + "epoch": 30.619808306709267, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 38336 + }, + { + "epoch": 30.620607028753994, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 38337 + }, + { + "epoch": 30.621405750798722, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 38338 + }, + { + "epoch": 30.62220447284345, + "grad_norm": 0.2333984375, + "learning_rate": 0.0001, + "loss": 1.0984, + "step": 38339 + }, + { + "epoch": 30.623003194888177, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 38340 + }, + { + "epoch": 30.62380191693291, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 38341 + }, + { + "epoch": 30.624600638977636, + "grad_norm": 0.1865234375, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 38342 + }, + { + "epoch": 30.625399361022364, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 38343 + }, + { + "epoch": 30.62619808306709, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 38344 + }, + { + "epoch": 30.626996805111823, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38345 + }, + { + "epoch": 30.62779552715655, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 38346 + }, + { + "epoch": 30.628594249201278, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 38347 + }, + { + "epoch": 30.629392971246006, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 38348 + }, + { + "epoch": 30.630191693290733, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 38349 + }, + { + "epoch": 30.630990415335464, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 38350 + }, + { + "epoch": 30.631789137380192, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 38351 + }, + { + "epoch": 30.63258785942492, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 38352 + }, + { + "epoch": 30.633386581469647, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 38353 + }, + { + "epoch": 30.63418530351438, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 38354 + }, + { + "epoch": 30.634984025559106, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 38355 + }, + { + "epoch": 30.635782747603834, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 38356 + }, + { + "epoch": 30.63658146964856, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 38357 + }, + { + "epoch": 30.63738019169329, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 38358 + }, + { + "epoch": 30.63817891373802, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 38359 + }, + { + "epoch": 30.638977635782748, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 38360 + }, + { + "epoch": 30.639776357827476, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 38361 + }, + { + "epoch": 30.640575079872203, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 38362 + }, + { + "epoch": 30.641373801916934, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38363 + }, + { + "epoch": 30.642172523961662, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 38364 + }, + { + "epoch": 30.64297124600639, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 38365 + }, + { + "epoch": 30.643769968051117, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 38366 + }, + { + "epoch": 30.644568690095845, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0754, + "step": 38367 + }, + { + "epoch": 30.645367412140576, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 38368 + }, + { + "epoch": 30.646166134185304, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 38369 + }, + { + "epoch": 30.64696485623003, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 38370 + }, + { + "epoch": 30.64776357827476, + "grad_norm": 0.19921875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 38371 + }, + { + "epoch": 30.64856230031949, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 38372 + }, + { + "epoch": 30.649361022364218, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 38373 + }, + { + "epoch": 30.650159744408946, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 38374 + }, + { + "epoch": 30.650958466453673, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 38375 + }, + { + "epoch": 30.6517571884984, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 38376 + }, + { + "epoch": 30.652555910543132, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 38377 + }, + { + "epoch": 30.65335463258786, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 38378 + }, + { + "epoch": 30.654153354632587, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 38379 + }, + { + "epoch": 30.654952076677315, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 38380 + }, + { + "epoch": 30.655750798722046, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0787, + "step": 38381 + }, + { + "epoch": 30.656549520766774, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 38382 + }, + { + "epoch": 30.6573482428115, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38383 + }, + { + "epoch": 30.65814696485623, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 38384 + }, + { + "epoch": 30.658945686900957, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 38385 + }, + { + "epoch": 30.659744408945688, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 38386 + }, + { + "epoch": 30.660543130990416, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 38387 + }, + { + "epoch": 30.661341853035143, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 38388 + }, + { + "epoch": 30.66214057507987, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 38389 + }, + { + "epoch": 30.662939297124602, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 38390 + }, + { + "epoch": 30.66373801916933, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 38391 + }, + { + "epoch": 30.664536741214057, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 38392 + }, + { + "epoch": 30.665335463258785, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 38393 + }, + { + "epoch": 30.666134185303516, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 38394 + }, + { + "epoch": 30.666932907348244, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 38395 + }, + { + "epoch": 30.66773162939297, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38396 + }, + { + "epoch": 30.6685303514377, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 38397 + }, + { + "epoch": 30.669329073482427, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 38398 + }, + { + "epoch": 30.670127795527158, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 38399 + }, + { + "epoch": 30.670926517571885, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 38400 + }, + { + "epoch": 30.671725239616613, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 38401 + }, + { + "epoch": 30.67252396166134, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 38402 + }, + { + "epoch": 30.673322683706072, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 38403 + }, + { + "epoch": 30.6741214057508, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 38404 + }, + { + "epoch": 30.674920127795527, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 38405 + }, + { + "epoch": 30.675718849840255, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 38406 + }, + { + "epoch": 30.676517571884983, + "grad_norm": 0.23828125, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 38407 + }, + { + "epoch": 30.677316293929714, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 38408 + }, + { + "epoch": 30.67811501597444, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 38409 + }, + { + "epoch": 30.67891373801917, + "grad_norm": 0.205078125, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 38410 + }, + { + "epoch": 30.679712460063897, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 38411 + }, + { + "epoch": 30.680511182108628, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 38412 + }, + { + "epoch": 30.681309904153355, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0767, + "step": 38413 + }, + { + "epoch": 30.682108626198083, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 38414 + }, + { + "epoch": 30.68290734824281, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 38415 + }, + { + "epoch": 30.68370607028754, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 38416 + }, + { + "epoch": 30.68450479233227, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 38417 + }, + { + "epoch": 30.685303514376997, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 38418 + }, + { + "epoch": 30.686102236421725, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 38419 + }, + { + "epoch": 30.686900958466452, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 38420 + }, + { + "epoch": 30.687699680511184, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 38421 + }, + { + "epoch": 30.68849840255591, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 38422 + }, + { + "epoch": 30.68929712460064, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 38423 + }, + { + "epoch": 30.690095846645367, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38424 + }, + { + "epoch": 30.690894568690094, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 38425 + }, + { + "epoch": 30.691693290734825, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0785, + "step": 38426 + }, + { + "epoch": 30.692492012779553, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 38427 + }, + { + "epoch": 30.69329073482428, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 38428 + }, + { + "epoch": 30.69408945686901, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 38429 + }, + { + "epoch": 30.69488817891374, + "grad_norm": 0.09716796875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 38430 + }, + { + "epoch": 30.695686900958467, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 38431 + }, + { + "epoch": 30.696485623003195, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 38432 + }, + { + "epoch": 30.697284345047922, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 38433 + }, + { + "epoch": 30.69808306709265, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 38434 + }, + { + "epoch": 30.69888178913738, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 38435 + }, + { + "epoch": 30.69968051118211, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0707, + "step": 38436 + }, + { + "epoch": 30.700479233226837, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 38437 + }, + { + "epoch": 30.701277955271564, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 38438 + }, + { + "epoch": 30.702076677316295, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 38439 + }, + { + "epoch": 30.702875399361023, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 38440 + }, + { + "epoch": 30.70367412140575, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 38441 + }, + { + "epoch": 30.70447284345048, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 38442 + }, + { + "epoch": 30.705271565495206, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 38443 + }, + { + "epoch": 30.706070287539937, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 38444 + }, + { + "epoch": 30.706869009584665, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 38445 + }, + { + "epoch": 30.707667731629392, + "grad_norm": 0.1025390625, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 38446 + }, + { + "epoch": 30.70846645367412, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 38447 + }, + { + "epoch": 30.70926517571885, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 38448 + }, + { + "epoch": 30.71006389776358, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 38449 + }, + { + "epoch": 30.710862619808307, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 38450 + }, + { + "epoch": 30.711661341853034, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 38451 + }, + { + "epoch": 30.712460063897765, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 38452 + }, + { + "epoch": 30.713258785942493, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 38453 + }, + { + "epoch": 30.71405750798722, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 38454 + }, + { + "epoch": 30.71485623003195, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 38455 + }, + { + "epoch": 30.715654952076676, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 38456 + }, + { + "epoch": 30.716453674121407, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 38457 + }, + { + "epoch": 30.717252396166135, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 38458 + }, + { + "epoch": 30.718051118210862, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 38459 + }, + { + "epoch": 30.71884984025559, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 38460 + }, + { + "epoch": 30.71964856230032, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 38461 + }, + { + "epoch": 30.72044728434505, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 38462 + }, + { + "epoch": 30.721246006389777, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 38463 + }, + { + "epoch": 30.722044728434504, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 38464 + }, + { + "epoch": 30.722843450479232, + "grad_norm": 0.10205078125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 38465 + }, + { + "epoch": 30.723642172523963, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 38466 + }, + { + "epoch": 30.72444089456869, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 38467 + }, + { + "epoch": 30.72523961661342, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0987, + "step": 38468 + }, + { + "epoch": 30.726038338658146, + "grad_norm": 0.09765625, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 38469 + }, + { + "epoch": 30.726837060702877, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 38470 + }, + { + "epoch": 30.727635782747605, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 38471 + }, + { + "epoch": 30.728434504792332, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38472 + }, + { + "epoch": 30.72923322683706, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 38473 + }, + { + "epoch": 30.730031948881788, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 38474 + }, + { + "epoch": 30.73083067092652, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 38475 + }, + { + "epoch": 30.731629392971247, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 38476 + }, + { + "epoch": 30.732428115015974, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 38477 + }, + { + "epoch": 30.733226837060702, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0757, + "step": 38478 + }, + { + "epoch": 30.734025559105433, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 38479 + }, + { + "epoch": 30.73482428115016, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 38480 + }, + { + "epoch": 30.73562300319489, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.107, + "step": 38481 + }, + { + "epoch": 30.736421725239616, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 38482 + }, + { + "epoch": 30.737220447284344, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.102, + "step": 38483 + }, + { + "epoch": 30.738019169329075, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38484 + }, + { + "epoch": 30.738817891373802, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0743, + "step": 38485 + }, + { + "epoch": 30.73961661341853, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 38486 + }, + { + "epoch": 30.740415335463258, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 38487 + }, + { + "epoch": 30.74121405750799, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 38488 + }, + { + "epoch": 30.742012779552716, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 38489 + }, + { + "epoch": 30.742811501597444, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 38490 + }, + { + "epoch": 30.74361022364217, + "grad_norm": 0.2890625, + "learning_rate": 0.0001, + "loss": 1.0777, + "step": 38491 + }, + { + "epoch": 30.7444089456869, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 38492 + }, + { + "epoch": 30.74520766773163, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 38493 + }, + { + "epoch": 30.74600638977636, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 38494 + }, + { + "epoch": 30.746805111821086, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 38495 + }, + { + "epoch": 30.747603833865814, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 38496 + }, + { + "epoch": 30.748402555910545, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 38497 + }, + { + "epoch": 30.749201277955272, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.1006, + "step": 38498 + }, + { + "epoch": 30.75, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 38499 + }, + { + "epoch": 30.750798722044728, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 38500 + }, + { + "epoch": 30.751597444089455, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 38501 + }, + { + "epoch": 30.752396166134186, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 38502 + }, + { + "epoch": 30.753194888178914, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 38503 + }, + { + "epoch": 30.75399361022364, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 38504 + }, + { + "epoch": 30.75479233226837, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 38505 + }, + { + "epoch": 30.7555910543131, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 38506 + }, + { + "epoch": 30.75638977635783, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 38507 + }, + { + "epoch": 30.757188498402556, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 38508 + }, + { + "epoch": 30.757987220447284, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 38509 + }, + { + "epoch": 30.75878594249201, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 38510 + }, + { + "epoch": 30.759584664536742, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 38511 + }, + { + "epoch": 30.76038338658147, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 38512 + }, + { + "epoch": 30.761182108626198, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.094, + "step": 38513 + }, + { + "epoch": 30.761980830670925, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 38514 + }, + { + "epoch": 30.762779552715656, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38515 + }, + { + "epoch": 30.763578274760384, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 38516 + }, + { + "epoch": 30.76437699680511, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 38517 + }, + { + "epoch": 30.76517571884984, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 38518 + }, + { + "epoch": 30.765974440894567, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 38519 + }, + { + "epoch": 30.766773162939298, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 38520 + }, + { + "epoch": 30.767571884984026, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 38521 + }, + { + "epoch": 30.768370607028753, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 38522 + }, + { + "epoch": 30.76916932907348, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0988, + "step": 38523 + }, + { + "epoch": 30.769968051118212, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 38524 + }, + { + "epoch": 30.77076677316294, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 38525 + }, + { + "epoch": 30.771565495207668, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38526 + }, + { + "epoch": 30.772364217252395, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 38527 + }, + { + "epoch": 30.773162939297123, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0991, + "step": 38528 + }, + { + "epoch": 30.773961661341854, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 38529 + }, + { + "epoch": 30.77476038338658, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 38530 + }, + { + "epoch": 30.77555910543131, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 38531 + }, + { + "epoch": 30.776357827476037, + "grad_norm": 0.2470703125, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 38532 + }, + { + "epoch": 30.777156549520768, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 38533 + }, + { + "epoch": 30.777955271565496, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 38534 + }, + { + "epoch": 30.778753993610223, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 38535 + }, + { + "epoch": 30.77955271565495, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 38536 + }, + { + "epoch": 30.78035143769968, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 38537 + }, + { + "epoch": 30.78115015974441, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 38538 + }, + { + "epoch": 30.781948881789138, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 38539 + }, + { + "epoch": 30.782747603833865, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 38540 + }, + { + "epoch": 30.783546325878593, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 38541 + }, + { + "epoch": 30.784345047923324, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 38542 + }, + { + "epoch": 30.78514376996805, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 38543 + }, + { + "epoch": 30.78594249201278, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 38544 + }, + { + "epoch": 30.786741214057507, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 38545 + }, + { + "epoch": 30.787539936102235, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 38546 + }, + { + "epoch": 30.788338658146966, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 38547 + }, + { + "epoch": 30.789137380191693, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 38548 + }, + { + "epoch": 30.78993610223642, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 38549 + }, + { + "epoch": 30.79073482428115, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 38550 + }, + { + "epoch": 30.79153354632588, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 38551 + }, + { + "epoch": 30.792332268370608, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 38552 + }, + { + "epoch": 30.793130990415335, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38553 + }, + { + "epoch": 30.793929712460063, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 38554 + }, + { + "epoch": 30.794728434504794, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 38555 + }, + { + "epoch": 30.79552715654952, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 38556 + }, + { + "epoch": 30.79632587859425, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 38557 + }, + { + "epoch": 30.797124600638977, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 38558 + }, + { + "epoch": 30.797923322683705, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 38559 + }, + { + "epoch": 30.798722044728436, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 38560 + }, + { + "epoch": 30.799520766773163, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38561 + }, + { + "epoch": 30.80031948881789, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 38562 + }, + { + "epoch": 30.80111821086262, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 38563 + }, + { + "epoch": 30.80191693290735, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38564 + }, + { + "epoch": 30.802715654952078, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 38565 + }, + { + "epoch": 30.803514376996805, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38566 + }, + { + "epoch": 30.804313099041533, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 38567 + }, + { + "epoch": 30.80511182108626, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 38568 + }, + { + "epoch": 30.80591054313099, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38569 + }, + { + "epoch": 30.80670926517572, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 38570 + }, + { + "epoch": 30.807507987220447, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 38571 + }, + { + "epoch": 30.808306709265175, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 38572 + }, + { + "epoch": 30.809105431309906, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 38573 + }, + { + "epoch": 30.809904153354633, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 38574 + }, + { + "epoch": 30.81070287539936, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 38575 + }, + { + "epoch": 30.81150159744409, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 38576 + }, + { + "epoch": 30.812300319488816, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 38577 + }, + { + "epoch": 30.813099041533548, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 38578 + }, + { + "epoch": 30.813897763578275, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 38579 + }, + { + "epoch": 30.814696485623003, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 38580 + }, + { + "epoch": 30.81549520766773, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 38581 + }, + { + "epoch": 30.81629392971246, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 38582 + }, + { + "epoch": 30.81709265175719, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 38583 + }, + { + "epoch": 30.817891373801917, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 38584 + }, + { + "epoch": 30.818690095846645, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 38585 + }, + { + "epoch": 30.819488817891372, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 38586 + }, + { + "epoch": 30.820287539936103, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 38587 + }, + { + "epoch": 30.82108626198083, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 38588 + }, + { + "epoch": 30.82188498402556, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 38589 + }, + { + "epoch": 30.822683706070286, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 38590 + }, + { + "epoch": 30.823482428115017, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 38591 + }, + { + "epoch": 30.824281150159745, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 38592 + }, + { + "epoch": 30.825079872204473, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 38593 + }, + { + "epoch": 30.8258785942492, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 38594 + }, + { + "epoch": 30.826677316293928, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 38595 + }, + { + "epoch": 30.82747603833866, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 38596 + }, + { + "epoch": 30.828274760383387, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 38597 + }, + { + "epoch": 30.829073482428115, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 38598 + }, + { + "epoch": 30.829872204472842, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0995, + "step": 38599 + }, + { + "epoch": 30.830670926517573, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 38600 + }, + { + "epoch": 30.8314696485623, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 38601 + }, + { + "epoch": 30.83226837060703, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 38602 + }, + { + "epoch": 30.833067092651756, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0949, + "step": 38603 + }, + { + "epoch": 30.833865814696484, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 38604 + }, + { + "epoch": 30.834664536741215, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0996, + "step": 38605 + }, + { + "epoch": 30.835463258785943, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 38606 + }, + { + "epoch": 30.83626198083067, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38607 + }, + { + "epoch": 30.837060702875398, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 38608 + }, + { + "epoch": 30.83785942492013, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 38609 + }, + { + "epoch": 30.838658146964857, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 38610 + }, + { + "epoch": 30.839456869009584, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 38611 + }, + { + "epoch": 30.840255591054312, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 38612 + }, + { + "epoch": 30.841054313099043, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 38613 + }, + { + "epoch": 30.84185303514377, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 38614 + }, + { + "epoch": 30.8426517571885, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.1044, + "step": 38615 + }, + { + "epoch": 30.843450479233226, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 38616 + }, + { + "epoch": 30.844249201277954, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38617 + }, + { + "epoch": 30.845047923322685, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 38618 + }, + { + "epoch": 30.845846645367413, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38619 + }, + { + "epoch": 30.84664536741214, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 38620 + }, + { + "epoch": 30.847444089456868, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 38621 + }, + { + "epoch": 30.8482428115016, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 38622 + }, + { + "epoch": 30.849041533546327, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 38623 + }, + { + "epoch": 30.849840255591054, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 38624 + }, + { + "epoch": 30.850638977635782, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 38625 + }, + { + "epoch": 30.85143769968051, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 38626 + }, + { + "epoch": 30.85223642172524, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 38627 + }, + { + "epoch": 30.85303514376997, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 38628 + }, + { + "epoch": 30.853833865814696, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 38629 + }, + { + "epoch": 30.854632587859424, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 38630 + }, + { + "epoch": 30.855431309904155, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 38631 + }, + { + "epoch": 30.856230031948883, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 38632 + }, + { + "epoch": 30.85702875399361, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 38633 + }, + { + "epoch": 30.857827476038338, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 38634 + }, + { + "epoch": 30.858626198083066, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 38635 + }, + { + "epoch": 30.859424920127797, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 38636 + }, + { + "epoch": 30.860223642172524, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38637 + }, + { + "epoch": 30.861022364217252, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0947, + "step": 38638 + }, + { + "epoch": 30.86182108626198, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 38639 + }, + { + "epoch": 30.86261980830671, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 38640 + }, + { + "epoch": 30.86341853035144, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 38641 + }, + { + "epoch": 30.864217252396166, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 38642 + }, + { + "epoch": 30.865015974440894, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 38643 + }, + { + "epoch": 30.86581469648562, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 38644 + }, + { + "epoch": 30.866613418530353, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 38645 + }, + { + "epoch": 30.86741214057508, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 38646 + }, + { + "epoch": 30.868210862619808, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 38647 + }, + { + "epoch": 30.869009584664536, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 38648 + }, + { + "epoch": 30.869808306709267, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 38649 + }, + { + "epoch": 30.870607028753994, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 38650 + }, + { + "epoch": 30.871405750798722, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 38651 + }, + { + "epoch": 30.87220447284345, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 38652 + }, + { + "epoch": 30.873003194888177, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0764, + "step": 38653 + }, + { + "epoch": 30.87380191693291, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 38654 + }, + { + "epoch": 30.874600638977636, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0748, + "step": 38655 + }, + { + "epoch": 30.875399361022364, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 38656 + }, + { + "epoch": 30.87619808306709, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 38657 + }, + { + "epoch": 30.876996805111823, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 38658 + }, + { + "epoch": 30.87779552715655, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 38659 + }, + { + "epoch": 30.878594249201278, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38660 + }, + { + "epoch": 30.879392971246006, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 38661 + }, + { + "epoch": 30.880191693290733, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 38662 + }, + { + "epoch": 30.880990415335464, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 38663 + }, + { + "epoch": 30.881789137380192, + "grad_norm": 0.27734375, + "learning_rate": 0.0001, + "loss": 1.0969, + "step": 38664 + }, + { + "epoch": 30.88258785942492, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 38665 + }, + { + "epoch": 30.883386581469647, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 38666 + }, + { + "epoch": 30.88418530351438, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 38667 + }, + { + "epoch": 30.884984025559106, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 38668 + }, + { + "epoch": 30.885782747603834, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0778, + "step": 38669 + }, + { + "epoch": 30.88658146964856, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 38670 + }, + { + "epoch": 30.88738019169329, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 38671 + }, + { + "epoch": 30.88817891373802, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 38672 + }, + { + "epoch": 30.888977635782748, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 38673 + }, + { + "epoch": 30.889776357827476, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 38674 + }, + { + "epoch": 30.890575079872203, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 38675 + }, + { + "epoch": 30.891373801916934, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 38676 + }, + { + "epoch": 30.892172523961662, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 38677 + }, + { + "epoch": 30.89297124600639, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 38678 + }, + { + "epoch": 30.893769968051117, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 38679 + }, + { + "epoch": 30.894568690095845, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38680 + }, + { + "epoch": 30.895367412140576, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 38681 + }, + { + "epoch": 30.896166134185304, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 38682 + }, + { + "epoch": 30.89696485623003, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 38683 + }, + { + "epoch": 30.89776357827476, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 38684 + }, + { + "epoch": 30.89856230031949, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 38685 + }, + { + "epoch": 30.899361022364218, + "grad_norm": 0.20703125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 38686 + }, + { + "epoch": 30.900159744408946, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 38687 + }, + { + "epoch": 30.900958466453673, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 38688 + }, + { + "epoch": 30.9017571884984, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 38689 + }, + { + "epoch": 30.902555910543132, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38690 + }, + { + "epoch": 30.90335463258786, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 38691 + }, + { + "epoch": 30.904153354632587, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 38692 + }, + { + "epoch": 30.904952076677315, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 38693 + }, + { + "epoch": 30.905750798722046, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 38694 + }, + { + "epoch": 30.906549520766774, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0792, + "step": 38695 + }, + { + "epoch": 30.9073482428115, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 38696 + }, + { + "epoch": 30.90814696485623, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 38697 + }, + { + "epoch": 30.908945686900957, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 38698 + }, + { + "epoch": 30.909744408945688, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 38699 + }, + { + "epoch": 30.910543130990416, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38700 + }, + { + "epoch": 30.911341853035143, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 38701 + }, + { + "epoch": 30.91214057507987, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 38702 + }, + { + "epoch": 30.912939297124602, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 38703 + }, + { + "epoch": 30.91373801916933, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 38704 + }, + { + "epoch": 30.914536741214057, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 38705 + }, + { + "epoch": 30.915335463258785, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 38706 + }, + { + "epoch": 30.916134185303516, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 38707 + }, + { + "epoch": 30.916932907348244, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 38708 + }, + { + "epoch": 30.91773162939297, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 38709 + }, + { + "epoch": 30.9185303514377, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.1014, + "step": 38710 + }, + { + "epoch": 30.919329073482427, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 38711 + }, + { + "epoch": 30.920127795527158, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 38712 + }, + { + "epoch": 30.920926517571885, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 38713 + }, + { + "epoch": 30.921725239616613, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 38714 + }, + { + "epoch": 30.92252396166134, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 38715 + }, + { + "epoch": 30.923322683706072, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 38716 + }, + { + "epoch": 30.9241214057508, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 38717 + }, + { + "epoch": 30.924920127795527, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 38718 + }, + { + "epoch": 30.925718849840255, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 38719 + }, + { + "epoch": 30.926517571884983, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 38720 + }, + { + "epoch": 30.927316293929714, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 38721 + }, + { + "epoch": 30.92811501597444, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 38722 + }, + { + "epoch": 30.92891373801917, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 38723 + }, + { + "epoch": 30.929712460063897, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 38724 + }, + { + "epoch": 30.930511182108628, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0967, + "step": 38725 + }, + { + "epoch": 30.931309904153355, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 38726 + }, + { + "epoch": 30.932108626198083, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 38727 + }, + { + "epoch": 30.93290734824281, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 38728 + }, + { + "epoch": 30.93370607028754, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 38729 + }, + { + "epoch": 30.93450479233227, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38730 + }, + { + "epoch": 30.935303514376997, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 38731 + }, + { + "epoch": 30.936102236421725, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 38732 + }, + { + "epoch": 30.936900958466452, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0941, + "step": 38733 + }, + { + "epoch": 30.937699680511184, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 38734 + }, + { + "epoch": 30.93849840255591, + "grad_norm": 0.21484375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38735 + }, + { + "epoch": 30.93929712460064, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 38736 + }, + { + "epoch": 30.940095846645367, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 38737 + }, + { + "epoch": 30.940894568690094, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 38738 + }, + { + "epoch": 30.941693290734825, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 38739 + }, + { + "epoch": 30.942492012779553, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0714, + "step": 38740 + }, + { + "epoch": 30.94329073482428, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 38741 + }, + { + "epoch": 30.94408945686901, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 38742 + }, + { + "epoch": 30.94488817891374, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 38743 + }, + { + "epoch": 30.945686900958467, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 38744 + }, + { + "epoch": 30.946485623003195, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 38745 + }, + { + "epoch": 30.947284345047922, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 38746 + }, + { + "epoch": 30.94808306709265, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 38747 + }, + { + "epoch": 30.94888178913738, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 38748 + }, + { + "epoch": 30.94968051118211, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 38749 + }, + { + "epoch": 30.950479233226837, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0752, + "step": 38750 + }, + { + "epoch": 30.951277955271564, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 38751 + }, + { + "epoch": 30.952076677316295, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 38752 + }, + { + "epoch": 30.952875399361023, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38753 + }, + { + "epoch": 30.95367412140575, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 38754 + }, + { + "epoch": 30.95447284345048, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 38755 + }, + { + "epoch": 30.955271565495206, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 38756 + }, + { + "epoch": 30.956070287539937, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 38757 + }, + { + "epoch": 30.956869009584665, + "grad_norm": 0.11083984375, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 38758 + }, + { + "epoch": 30.957667731629392, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 38759 + }, + { + "epoch": 30.95846645367412, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 38760 + }, + { + "epoch": 30.95926517571885, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0981, + "step": 38761 + }, + { + "epoch": 30.96006389776358, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 38762 + }, + { + "epoch": 30.960862619808307, + "grad_norm": 0.1943359375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 38763 + }, + { + "epoch": 30.961661341853034, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 38764 + }, + { + "epoch": 30.962460063897765, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0773, + "step": 38765 + }, + { + "epoch": 30.963258785942493, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 38766 + }, + { + "epoch": 30.96405750798722, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0762, + "step": 38767 + }, + { + "epoch": 30.96485623003195, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 38768 + }, + { + "epoch": 30.965654952076676, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 38769 + }, + { + "epoch": 30.966453674121407, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 38770 + }, + { + "epoch": 30.967252396166135, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 38771 + }, + { + "epoch": 30.968051118210862, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 38772 + }, + { + "epoch": 30.96884984025559, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 38773 + }, + { + "epoch": 30.96964856230032, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 38774 + }, + { + "epoch": 30.97044728434505, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 38775 + }, + { + "epoch": 30.971246006389777, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.077, + "step": 38776 + }, + { + "epoch": 30.972044728434504, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 38777 + }, + { + "epoch": 30.972843450479232, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0953, + "step": 38778 + }, + { + "epoch": 30.973642172523963, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 38779 + }, + { + "epoch": 30.97444089456869, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 38780 + }, + { + "epoch": 30.97523961661342, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 38781 + }, + { + "epoch": 30.976038338658146, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 38782 + }, + { + "epoch": 30.976837060702877, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 38783 + }, + { + "epoch": 30.977635782747605, + "grad_norm": 0.2265625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 38784 + }, + { + "epoch": 30.978434504792332, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38785 + }, + { + "epoch": 30.97923322683706, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38786 + }, + { + "epoch": 30.980031948881788, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 38787 + }, + { + "epoch": 30.98083067092652, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 38788 + }, + { + "epoch": 30.981629392971247, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 38789 + }, + { + "epoch": 30.982428115015974, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 38790 + }, + { + "epoch": 30.983226837060702, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 38791 + }, + { + "epoch": 30.984025559105433, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 38792 + }, + { + "epoch": 30.98482428115016, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 38793 + }, + { + "epoch": 30.98562300319489, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38794 + }, + { + "epoch": 30.986421725239616, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 38795 + }, + { + "epoch": 30.987220447284344, + "grad_norm": 0.248046875, + "learning_rate": 0.0001, + "loss": 1.1023, + "step": 38796 + }, + { + "epoch": 30.988019169329075, + "grad_norm": 0.2275390625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 38797 + }, + { + "epoch": 30.988817891373802, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 38798 + }, + { + "epoch": 30.98961661341853, + "grad_norm": 0.212890625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 38799 + }, + { + "epoch": 30.990415335463258, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 38800 + }, + { + "epoch": 30.99121405750799, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 38801 + }, + { + "epoch": 30.992012779552716, + "grad_norm": 0.177734375, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 38802 + }, + { + "epoch": 30.992811501597444, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 38803 + }, + { + "epoch": 30.99361022364217, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 38804 + }, + { + "epoch": 30.9944089456869, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 38805 + }, + { + "epoch": 30.99520766773163, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 38806 + }, + { + "epoch": 30.99600638977636, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 38807 + }, + { + "epoch": 30.996805111821086, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 38808 + }, + { + "epoch": 30.997603833865814, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 38809 + }, + { + "epoch": 30.998402555910545, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 38810 + }, + { + "epoch": 30.999201277955272, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0762, + "step": 38811 + }, + { + "epoch": 31.0, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 38812 + }, + { + "epoch": 31.000798722044728, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 38813 + }, + { + "epoch": 31.001597444089455, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 38814 + }, + { + "epoch": 31.002396166134186, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38815 + }, + { + "epoch": 31.003194888178914, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0782, + "step": 38816 + }, + { + "epoch": 31.00399361022364, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 38817 + }, + { + "epoch": 31.00479233226837, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 38818 + }, + { + "epoch": 31.0055910543131, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 38819 + }, + { + "epoch": 31.00638977635783, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 38820 + }, + { + "epoch": 31.007188498402556, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 38821 + }, + { + "epoch": 31.007987220447284, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 38822 + }, + { + "epoch": 31.00878594249201, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.1053, + "step": 38823 + }, + { + "epoch": 31.009584664536742, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38824 + }, + { + "epoch": 31.01038338658147, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 38825 + }, + { + "epoch": 31.011182108626198, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 38826 + }, + { + "epoch": 31.011980830670925, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 38827 + }, + { + "epoch": 31.012779552715656, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 38828 + }, + { + "epoch": 31.013578274760384, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 38829 + }, + { + "epoch": 31.01437699680511, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 38830 + }, + { + "epoch": 31.01517571884984, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 38831 + }, + { + "epoch": 31.015974440894567, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38832 + }, + { + "epoch": 31.016773162939298, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 38833 + }, + { + "epoch": 31.017571884984026, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 38834 + }, + { + "epoch": 31.018370607028753, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0986, + "step": 38835 + }, + { + "epoch": 31.01916932907348, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 38836 + }, + { + "epoch": 31.019968051118212, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38837 + }, + { + "epoch": 31.02076677316294, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 38838 + }, + { + "epoch": 31.021565495207668, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 38839 + }, + { + "epoch": 31.022364217252395, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 38840 + }, + { + "epoch": 31.023162939297123, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 38841 + }, + { + "epoch": 31.023961661341854, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 38842 + }, + { + "epoch": 31.02476038338658, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 38843 + }, + { + "epoch": 31.02555910543131, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 38844 + }, + { + "epoch": 31.026357827476037, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 38845 + }, + { + "epoch": 31.027156549520768, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 38846 + }, + { + "epoch": 31.027955271565496, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 38847 + }, + { + "epoch": 31.028753993610223, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 38848 + }, + { + "epoch": 31.02955271565495, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 38849 + }, + { + "epoch": 31.03035143769968, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38850 + }, + { + "epoch": 31.03115015974441, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 38851 + }, + { + "epoch": 31.031948881789138, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 38852 + }, + { + "epoch": 31.032747603833865, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 38853 + }, + { + "epoch": 31.033546325878593, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 38854 + }, + { + "epoch": 31.034345047923324, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 38855 + }, + { + "epoch": 31.03514376996805, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 38856 + }, + { + "epoch": 31.03594249201278, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 38857 + }, + { + "epoch": 31.036741214057507, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 38858 + }, + { + "epoch": 31.037539936102238, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0819, + "step": 38859 + }, + { + "epoch": 31.038338658146966, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 38860 + }, + { + "epoch": 31.039137380191693, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 38861 + }, + { + "epoch": 31.03993610223642, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 38862 + }, + { + "epoch": 31.04073482428115, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 38863 + }, + { + "epoch": 31.04153354632588, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 38864 + }, + { + "epoch": 31.042332268370608, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 38865 + }, + { + "epoch": 31.043130990415335, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38866 + }, + { + "epoch": 31.043929712460063, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38867 + }, + { + "epoch": 31.044728434504794, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 38868 + }, + { + "epoch": 31.04552715654952, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38869 + }, + { + "epoch": 31.04632587859425, + "grad_norm": 0.267578125, + "learning_rate": 0.0001, + "loss": 1.0761, + "step": 38870 + }, + { + "epoch": 31.047124600638977, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 38871 + }, + { + "epoch": 31.047923322683705, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 38872 + }, + { + "epoch": 31.048722044728436, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 38873 + }, + { + "epoch": 31.049520766773163, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 38874 + }, + { + "epoch": 31.05031948881789, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 38875 + }, + { + "epoch": 31.05111821086262, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 38876 + }, + { + "epoch": 31.05191693290735, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 38877 + }, + { + "epoch": 31.052715654952078, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38878 + }, + { + "epoch": 31.053514376996805, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 38879 + }, + { + "epoch": 31.054313099041533, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 38880 + }, + { + "epoch": 31.05511182108626, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 38881 + }, + { + "epoch": 31.05591054313099, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 38882 + }, + { + "epoch": 31.05670926517572, + "grad_norm": 0.1806640625, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 38883 + }, + { + "epoch": 31.057507987220447, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 38884 + }, + { + "epoch": 31.058306709265175, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 38885 + }, + { + "epoch": 31.059105431309906, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 38886 + }, + { + "epoch": 31.059904153354633, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 38887 + }, + { + "epoch": 31.06070287539936, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 38888 + }, + { + "epoch": 31.06150159744409, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 38889 + }, + { + "epoch": 31.062300319488816, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 38890 + }, + { + "epoch": 31.063099041533548, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 38891 + }, + { + "epoch": 31.063897763578275, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 38892 + }, + { + "epoch": 31.064696485623003, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 38893 + }, + { + "epoch": 31.06549520766773, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 38894 + }, + { + "epoch": 31.06629392971246, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 38895 + }, + { + "epoch": 31.06709265175719, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 38896 + }, + { + "epoch": 31.067891373801917, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 38897 + }, + { + "epoch": 31.068690095846645, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 38898 + }, + { + "epoch": 31.069488817891372, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0762, + "step": 38899 + }, + { + "epoch": 31.070287539936103, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 38900 + }, + { + "epoch": 31.07108626198083, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0828, + "step": 38901 + }, + { + "epoch": 31.07188498402556, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 38902 + }, + { + "epoch": 31.072683706070286, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 38903 + }, + { + "epoch": 31.073482428115017, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 38904 + }, + { + "epoch": 31.074281150159745, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 38905 + }, + { + "epoch": 31.075079872204473, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 38906 + }, + { + "epoch": 31.0758785942492, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 38907 + }, + { + "epoch": 31.076677316293928, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 38908 + }, + { + "epoch": 31.07747603833866, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38909 + }, + { + "epoch": 31.078274760383387, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 38910 + }, + { + "epoch": 31.079073482428115, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 38911 + }, + { + "epoch": 31.079872204472842, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 38912 + }, + { + "epoch": 31.080670926517573, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.088, + "step": 38913 + }, + { + "epoch": 31.0814696485623, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 38914 + }, + { + "epoch": 31.08226837060703, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 38915 + }, + { + "epoch": 31.083067092651756, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 38916 + }, + { + "epoch": 31.083865814696484, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 38917 + }, + { + "epoch": 31.084664536741215, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 38918 + }, + { + "epoch": 31.085463258785943, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 38919 + }, + { + "epoch": 31.08626198083067, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 38920 + }, + { + "epoch": 31.087060702875398, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0796, + "step": 38921 + }, + { + "epoch": 31.08785942492013, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 38922 + }, + { + "epoch": 31.088658146964857, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 38923 + }, + { + "epoch": 31.089456869009584, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 38924 + }, + { + "epoch": 31.090255591054312, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 38925 + }, + { + "epoch": 31.09105431309904, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 38926 + }, + { + "epoch": 31.09185303514377, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 38927 + }, + { + "epoch": 31.0926517571885, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0779, + "step": 38928 + }, + { + "epoch": 31.093450479233226, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 38929 + }, + { + "epoch": 31.094249201277954, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0788, + "step": 38930 + }, + { + "epoch": 31.095047923322685, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38931 + }, + { + "epoch": 31.095846645367413, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38932 + }, + { + "epoch": 31.09664536741214, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 38933 + }, + { + "epoch": 31.097444089456868, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 38934 + }, + { + "epoch": 31.0982428115016, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 38935 + }, + { + "epoch": 31.099041533546327, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 38936 + }, + { + "epoch": 31.099840255591054, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 38937 + }, + { + "epoch": 31.100638977635782, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 38938 + }, + { + "epoch": 31.10143769968051, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0679, + "step": 38939 + }, + { + "epoch": 31.10223642172524, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 38940 + }, + { + "epoch": 31.10303514376997, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 38941 + }, + { + "epoch": 31.103833865814696, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 38942 + }, + { + "epoch": 31.104632587859424, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0964, + "step": 38943 + }, + { + "epoch": 31.105431309904155, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 38944 + }, + { + "epoch": 31.106230031948883, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 38945 + }, + { + "epoch": 31.10702875399361, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 38946 + }, + { + "epoch": 31.107827476038338, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 38947 + }, + { + "epoch": 31.108626198083066, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 38948 + }, + { + "epoch": 31.109424920127797, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 38949 + }, + { + "epoch": 31.110223642172524, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 38950 + }, + { + "epoch": 31.111022364217252, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 38951 + }, + { + "epoch": 31.11182108626198, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 38952 + }, + { + "epoch": 31.11261980830671, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 38953 + }, + { + "epoch": 31.11341853035144, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0802, + "step": 38954 + }, + { + "epoch": 31.114217252396166, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0765, + "step": 38955 + }, + { + "epoch": 31.115015974440894, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0966, + "step": 38956 + }, + { + "epoch": 31.11581469648562, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 38957 + }, + { + "epoch": 31.116613418530353, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 38958 + }, + { + "epoch": 31.11741214057508, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 38959 + }, + { + "epoch": 31.118210862619808, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 38960 + }, + { + "epoch": 31.119009584664536, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 38961 + }, + { + "epoch": 31.119808306709267, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0735, + "step": 38962 + }, + { + "epoch": 31.120607028753994, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0965, + "step": 38963 + }, + { + "epoch": 31.121405750798722, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 38964 + }, + { + "epoch": 31.12220447284345, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 38965 + }, + { + "epoch": 31.123003194888177, + "grad_norm": 0.12353515625, + "learning_rate": 0.0001, + "loss": 1.097, + "step": 38966 + }, + { + "epoch": 31.12380191693291, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 38967 + }, + { + "epoch": 31.124600638977636, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 38968 + }, + { + "epoch": 31.125399361022364, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 38969 + }, + { + "epoch": 31.12619808306709, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 38970 + }, + { + "epoch": 31.126996805111823, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 38971 + }, + { + "epoch": 31.12779552715655, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0902, + "step": 38972 + }, + { + "epoch": 31.128594249201278, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 38973 + }, + { + "epoch": 31.129392971246006, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 38974 + }, + { + "epoch": 31.130191693290733, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0731, + "step": 38975 + }, + { + "epoch": 31.130990415335464, + "grad_norm": 0.181640625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 38976 + }, + { + "epoch": 31.131789137380192, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 38977 + }, + { + "epoch": 31.13258785942492, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 38978 + }, + { + "epoch": 31.133386581469647, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 38979 + }, + { + "epoch": 31.13418530351438, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0935, + "step": 38980 + }, + { + "epoch": 31.134984025559106, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 38981 + }, + { + "epoch": 31.135782747603834, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 38982 + }, + { + "epoch": 31.13658146964856, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0847, + "step": 38983 + }, + { + "epoch": 31.13738019169329, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 38984 + }, + { + "epoch": 31.13817891373802, + "grad_norm": 0.09912109375, + "learning_rate": 0.0001, + "loss": 1.1034, + "step": 38985 + }, + { + "epoch": 31.138977635782748, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 38986 + }, + { + "epoch": 31.139776357827476, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 38987 + }, + { + "epoch": 31.140575079872203, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 38988 + }, + { + "epoch": 31.141373801916934, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 38989 + }, + { + "epoch": 31.142172523961662, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0786, + "step": 38990 + }, + { + "epoch": 31.14297124600639, + "grad_norm": 0.2060546875, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 38991 + }, + { + "epoch": 31.143769968051117, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 38992 + }, + { + "epoch": 31.144568690095845, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 38993 + }, + { + "epoch": 31.145367412140576, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 38994 + }, + { + "epoch": 31.146166134185304, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 38995 + }, + { + "epoch": 31.14696485623003, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 38996 + }, + { + "epoch": 31.14776357827476, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 38997 + }, + { + "epoch": 31.14856230031949, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0748, + "step": 38998 + }, + { + "epoch": 31.149361022364218, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 38999 + }, + { + "epoch": 31.150159744408946, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 39000 + }, + { + "epoch": 31.150958466453673, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 39001 + }, + { + "epoch": 31.1517571884984, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 39002 + }, + { + "epoch": 31.152555910543132, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.0813, + "step": 39003 + }, + { + "epoch": 31.15335463258786, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 39004 + }, + { + "epoch": 31.154153354632587, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0736, + "step": 39005 + }, + { + "epoch": 31.154952076677315, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0934, + "step": 39006 + }, + { + "epoch": 31.155750798722046, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 39007 + }, + { + "epoch": 31.156549520766774, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 39008 + }, + { + "epoch": 31.1573482428115, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 39009 + }, + { + "epoch": 31.15814696485623, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 39010 + }, + { + "epoch": 31.15894568690096, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 39011 + }, + { + "epoch": 31.159744408945688, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 39012 + }, + { + "epoch": 31.160543130990416, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 39013 + }, + { + "epoch": 31.161341853035143, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 39014 + }, + { + "epoch": 31.16214057507987, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 39015 + }, + { + "epoch": 31.162939297124602, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 39016 + }, + { + "epoch": 31.16373801916933, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 39017 + }, + { + "epoch": 31.164536741214057, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 39018 + }, + { + "epoch": 31.165335463258785, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 39019 + }, + { + "epoch": 31.166134185303516, + "grad_norm": 0.11181640625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 39020 + }, + { + "epoch": 31.166932907348244, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 39021 + }, + { + "epoch": 31.16773162939297, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 39022 + }, + { + "epoch": 31.1685303514377, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 39023 + }, + { + "epoch": 31.169329073482427, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 39024 + }, + { + "epoch": 31.170127795527158, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 39025 + }, + { + "epoch": 31.170926517571885, + "grad_norm": 0.2119140625, + "learning_rate": 0.0001, + "loss": 1.0928, + "step": 39026 + }, + { + "epoch": 31.171725239616613, + "grad_norm": 0.2158203125, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 39027 + }, + { + "epoch": 31.17252396166134, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 39028 + }, + { + "epoch": 31.173322683706072, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 39029 + }, + { + "epoch": 31.1741214057508, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 39030 + }, + { + "epoch": 31.174920127795527, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 39031 + }, + { + "epoch": 31.175718849840255, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 39032 + }, + { + "epoch": 31.176517571884983, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0945, + "step": 39033 + }, + { + "epoch": 31.177316293929714, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 39034 + }, + { + "epoch": 31.17811501597444, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 39035 + }, + { + "epoch": 31.17891373801917, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 39036 + }, + { + "epoch": 31.179712460063897, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 39037 + }, + { + "epoch": 31.180511182108628, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0804, + "step": 39038 + }, + { + "epoch": 31.181309904153355, + "grad_norm": 0.11865234375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 39039 + }, + { + "epoch": 31.182108626198083, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 39040 + }, + { + "epoch": 31.18290734824281, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 39041 + }, + { + "epoch": 31.18370607028754, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 39042 + }, + { + "epoch": 31.18450479233227, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 39043 + }, + { + "epoch": 31.185303514376997, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 39044 + }, + { + "epoch": 31.186102236421725, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 39045 + }, + { + "epoch": 31.186900958466452, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 39046 + }, + { + "epoch": 31.187699680511184, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 39047 + }, + { + "epoch": 31.18849840255591, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 39048 + }, + { + "epoch": 31.18929712460064, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 39049 + }, + { + "epoch": 31.190095846645367, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0765, + "step": 39050 + }, + { + "epoch": 31.190894568690094, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 39051 + }, + { + "epoch": 31.191693290734825, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 39052 + }, + { + "epoch": 31.192492012779553, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 39053 + }, + { + "epoch": 31.19329073482428, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 39054 + }, + { + "epoch": 31.19408945686901, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 39055 + }, + { + "epoch": 31.19488817891374, + "grad_norm": 0.2236328125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 39056 + }, + { + "epoch": 31.195686900958467, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 39057 + }, + { + "epoch": 31.196485623003195, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 39058 + }, + { + "epoch": 31.197284345047922, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 39059 + }, + { + "epoch": 31.19808306709265, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 39060 + }, + { + "epoch": 31.19888178913738, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0811, + "step": 39061 + }, + { + "epoch": 31.19968051118211, + "grad_norm": 0.2197265625, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 39062 + }, + { + "epoch": 31.200479233226837, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 39063 + }, + { + "epoch": 31.201277955271564, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 39064 + }, + { + "epoch": 31.202076677316295, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 39065 + }, + { + "epoch": 31.202875399361023, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0911, + "step": 39066 + }, + { + "epoch": 31.20367412140575, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 39067 + }, + { + "epoch": 31.20447284345048, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 39068 + }, + { + "epoch": 31.205271565495206, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0812, + "step": 39069 + }, + { + "epoch": 31.206070287539937, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 39070 + }, + { + "epoch": 31.206869009584665, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 39071 + }, + { + "epoch": 31.207667731629392, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 39072 + }, + { + "epoch": 31.20846645367412, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 39073 + }, + { + "epoch": 31.20926517571885, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 39074 + }, + { + "epoch": 31.21006389776358, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 39075 + }, + { + "epoch": 31.210862619808307, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 39076 + }, + { + "epoch": 31.211661341853034, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 39077 + }, + { + "epoch": 31.212460063897762, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 39078 + }, + { + "epoch": 31.213258785942493, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0874, + "step": 39079 + }, + { + "epoch": 31.21405750798722, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 39080 + }, + { + "epoch": 31.21485623003195, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 39081 + }, + { + "epoch": 31.215654952076676, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0919, + "step": 39082 + }, + { + "epoch": 31.216453674121407, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0979, + "step": 39083 + }, + { + "epoch": 31.217252396166135, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 39084 + }, + { + "epoch": 31.218051118210862, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 39085 + }, + { + "epoch": 31.21884984025559, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 39086 + }, + { + "epoch": 31.21964856230032, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.087, + "step": 39087 + }, + { + "epoch": 31.22044728434505, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 39088 + }, + { + "epoch": 31.221246006389777, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 39089 + }, + { + "epoch": 31.222044728434504, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 39090 + }, + { + "epoch": 31.222843450479232, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 39091 + }, + { + "epoch": 31.223642172523963, + "grad_norm": 0.2138671875, + "learning_rate": 0.0001, + "loss": 1.0821, + "step": 39092 + }, + { + "epoch": 31.22444089456869, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 39093 + }, + { + "epoch": 31.22523961661342, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 39094 + }, + { + "epoch": 31.226038338658146, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 39095 + }, + { + "epoch": 31.226837060702877, + "grad_norm": 0.232421875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 39096 + }, + { + "epoch": 31.227635782747605, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 39097 + }, + { + "epoch": 31.228434504792332, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0801, + "step": 39098 + }, + { + "epoch": 31.22923322683706, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 39099 + }, + { + "epoch": 31.230031948881788, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 39100 + }, + { + "epoch": 31.23083067092652, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0913, + "step": 39101 + }, + { + "epoch": 31.231629392971247, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 39102 + }, + { + "epoch": 31.232428115015974, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 39103 + }, + { + "epoch": 31.233226837060702, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0794, + "step": 39104 + }, + { + "epoch": 31.234025559105433, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 39105 + }, + { + "epoch": 31.23482428115016, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 39106 + }, + { + "epoch": 31.23562300319489, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 39107 + }, + { + "epoch": 31.236421725239616, + "grad_norm": 0.12158203125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 39108 + }, + { + "epoch": 31.237220447284344, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 39109 + }, + { + "epoch": 31.238019169329075, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0974, + "step": 39110 + }, + { + "epoch": 31.238817891373802, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 39111 + }, + { + "epoch": 31.23961661341853, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 39112 + }, + { + "epoch": 31.240415335463258, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 39113 + }, + { + "epoch": 31.24121405750799, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0759, + "step": 39114 + }, + { + "epoch": 31.242012779552716, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 39115 + }, + { + "epoch": 31.242811501597444, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0883, + "step": 39116 + }, + { + "epoch": 31.24361022364217, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 39117 + }, + { + "epoch": 31.2444089456869, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 39118 + }, + { + "epoch": 31.24520766773163, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 39119 + }, + { + "epoch": 31.24600638977636, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 39120 + }, + { + "epoch": 31.246805111821086, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 39121 + }, + { + "epoch": 31.247603833865814, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 39122 + }, + { + "epoch": 31.248402555910545, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 39123 + }, + { + "epoch": 31.249201277955272, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0946, + "step": 39124 + }, + { + "epoch": 31.25, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 39125 + }, + { + "epoch": 31.250798722044728, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 39126 + }, + { + "epoch": 31.251597444089455, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0748, + "step": 39127 + }, + { + "epoch": 31.252396166134186, + "grad_norm": 0.1162109375, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 39128 + }, + { + "epoch": 31.253194888178914, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0809, + "step": 39129 + }, + { + "epoch": 31.25399361022364, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0754, + "step": 39130 + }, + { + "epoch": 31.25479233226837, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 39131 + }, + { + "epoch": 31.2555910543131, + "grad_norm": 0.115234375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 39132 + }, + { + "epoch": 31.25638977635783, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 39133 + }, + { + "epoch": 31.257188498402556, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 39134 + }, + { + "epoch": 31.257987220447284, + "grad_norm": 0.10595703125, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 39135 + }, + { + "epoch": 31.25878594249201, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.1027, + "step": 39136 + }, + { + "epoch": 31.259584664536742, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.0898, + "step": 39137 + }, + { + "epoch": 31.26038338658147, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 39138 + }, + { + "epoch": 31.261182108626198, + "grad_norm": 0.10498046875, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 39139 + }, + { + "epoch": 31.261980830670925, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 39140 + }, + { + "epoch": 31.262779552715656, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 39141 + }, + { + "epoch": 31.263578274760384, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 39142 + }, + { + "epoch": 31.26437699680511, + "grad_norm": 0.11376953125, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 39143 + }, + { + "epoch": 31.26517571884984, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 39144 + }, + { + "epoch": 31.265974440894567, + "grad_norm": 0.1220703125, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 39145 + }, + { + "epoch": 31.266773162939298, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0959, + "step": 39146 + }, + { + "epoch": 31.267571884984026, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0713, + "step": 39147 + }, + { + "epoch": 31.268370607028753, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0888, + "step": 39148 + }, + { + "epoch": 31.26916932907348, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 39149 + }, + { + "epoch": 31.269968051118212, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 39150 + }, + { + "epoch": 31.27076677316294, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 39151 + }, + { + "epoch": 31.271565495207668, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 39152 + }, + { + "epoch": 31.272364217252395, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 39153 + }, + { + "epoch": 31.273162939297123, + "grad_norm": 0.1240234375, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 39154 + }, + { + "epoch": 31.273961661341854, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 39155 + }, + { + "epoch": 31.27476038338658, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 39156 + }, + { + "epoch": 31.27555910543131, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 39157 + }, + { + "epoch": 31.276357827476037, + "grad_norm": 0.10791015625, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 39158 + }, + { + "epoch": 31.277156549520768, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 39159 + }, + { + "epoch": 31.277955271565496, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0975, + "step": 39160 + }, + { + "epoch": 31.278753993610223, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 39161 + }, + { + "epoch": 31.27955271565495, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0758, + "step": 39162 + }, + { + "epoch": 31.28035143769968, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 39163 + }, + { + "epoch": 31.28115015974441, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0798, + "step": 39164 + }, + { + "epoch": 31.281948881789138, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 39165 + }, + { + "epoch": 31.282747603833865, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 39166 + }, + { + "epoch": 31.283546325878593, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 39167 + }, + { + "epoch": 31.284345047923324, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0799, + "step": 39168 + }, + { + "epoch": 31.28514376996805, + "grad_norm": 0.1884765625, + "learning_rate": 0.0001, + "loss": 1.0791, + "step": 39169 + }, + { + "epoch": 31.28594249201278, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 39170 + }, + { + "epoch": 31.286741214057507, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 39171 + }, + { + "epoch": 31.287539936102238, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 39172 + }, + { + "epoch": 31.288338658146966, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 39173 + }, + { + "epoch": 31.289137380191693, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 39174 + }, + { + "epoch": 31.28993610223642, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 39175 + }, + { + "epoch": 31.29073482428115, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 39176 + }, + { + "epoch": 31.29153354632588, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 39177 + }, + { + "epoch": 31.292332268370608, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 39178 + }, + { + "epoch": 31.293130990415335, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 39179 + }, + { + "epoch": 31.293929712460063, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 39180 + }, + { + "epoch": 31.294728434504794, + "grad_norm": 0.208984375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 39181 + }, + { + "epoch": 31.29552715654952, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 39182 + }, + { + "epoch": 31.29632587859425, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 39183 + }, + { + "epoch": 31.297124600638977, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 39184 + }, + { + "epoch": 31.297923322683705, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0818, + "step": 39185 + }, + { + "epoch": 31.298722044728436, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 39186 + }, + { + "epoch": 31.299520766773163, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.093, + "step": 39187 + }, + { + "epoch": 31.30031948881789, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 39188 + }, + { + "epoch": 31.30111821086262, + "grad_norm": 0.236328125, + "learning_rate": 0.0001, + "loss": 1.08, + "step": 39189 + }, + { + "epoch": 31.30191693290735, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 39190 + }, + { + "epoch": 31.302715654952078, + "grad_norm": 0.25390625, + "learning_rate": 0.0001, + "loss": 1.0845, + "step": 39191 + }, + { + "epoch": 31.303514376996805, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 39192 + }, + { + "epoch": 31.304313099041533, + "grad_norm": 0.24609375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 39193 + }, + { + "epoch": 31.30511182108626, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 39194 + }, + { + "epoch": 31.30591054313099, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 39195 + }, + { + "epoch": 31.30670926517572, + "grad_norm": 0.10693359375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 39196 + }, + { + "epoch": 31.307507987220447, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 39197 + }, + { + "epoch": 31.308306709265175, + "grad_norm": 0.1640625, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 39198 + }, + { + "epoch": 31.309105431309906, + "grad_norm": 0.1083984375, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 39199 + }, + { + "epoch": 31.309904153354633, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 39200 + }, + { + "epoch": 31.31070287539936, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0781, + "step": 39201 + }, + { + "epoch": 31.31150159744409, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 39202 + }, + { + "epoch": 31.312300319488816, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 39203 + }, + { + "epoch": 31.313099041533548, + "grad_norm": 0.1591796875, + "learning_rate": 0.0001, + "loss": 1.0954, + "step": 39204 + }, + { + "epoch": 31.313897763578275, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 39205 + }, + { + "epoch": 31.314696485623003, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 39206 + }, + { + "epoch": 31.31549520766773, + "grad_norm": 0.1669921875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 39207 + }, + { + "epoch": 31.31629392971246, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 39208 + }, + { + "epoch": 31.31709265175719, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 39209 + }, + { + "epoch": 31.317891373801917, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 39210 + }, + { + "epoch": 31.318690095846645, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0853, + "step": 39211 + }, + { + "epoch": 31.319488817891372, + "grad_norm": 0.111328125, + "learning_rate": 0.0001, + "loss": 1.0768, + "step": 39212 + }, + { + "epoch": 31.320287539936103, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 39213 + }, + { + "epoch": 31.32108626198083, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 39214 + }, + { + "epoch": 31.32188498402556, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 39215 + }, + { + "epoch": 31.322683706070286, + "grad_norm": 0.1904296875, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 39216 + }, + { + "epoch": 31.323482428115017, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 39217 + }, + { + "epoch": 31.324281150159745, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 39218 + }, + { + "epoch": 31.325079872204473, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 39219 + }, + { + "epoch": 31.3258785942492, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 39220 + }, + { + "epoch": 31.326677316293928, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 39221 + }, + { + "epoch": 31.32747603833866, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0971, + "step": 39222 + }, + { + "epoch": 31.328274760383387, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0842, + "step": 39223 + }, + { + "epoch": 31.329073482428115, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 39224 + }, + { + "epoch": 31.329872204472842, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 39225 + }, + { + "epoch": 31.330670926517573, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 39226 + }, + { + "epoch": 31.3314696485623, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 39227 + }, + { + "epoch": 31.33226837060703, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0789, + "step": 39228 + }, + { + "epoch": 31.333067092651756, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 39229 + }, + { + "epoch": 31.333865814696484, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 39230 + }, + { + "epoch": 31.334664536741215, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 39231 + }, + { + "epoch": 31.335463258785943, + "grad_norm": 0.103515625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 39232 + }, + { + "epoch": 31.33626198083067, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 39233 + }, + { + "epoch": 31.337060702875398, + "grad_norm": 0.1103515625, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 39234 + }, + { + "epoch": 31.33785942492013, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0848, + "step": 39235 + }, + { + "epoch": 31.338658146964857, + "grad_norm": 0.11572265625, + "learning_rate": 0.0001, + "loss": 1.0751, + "step": 39236 + }, + { + "epoch": 31.339456869009584, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 39237 + }, + { + "epoch": 31.340255591054312, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 39238 + }, + { + "epoch": 31.341054313099043, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 39239 + }, + { + "epoch": 31.34185303514377, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 39240 + }, + { + "epoch": 31.3426517571885, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 39241 + }, + { + "epoch": 31.343450479233226, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 39242 + }, + { + "epoch": 31.344249201277954, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 39243 + }, + { + "epoch": 31.345047923322685, + "grad_norm": 0.234375, + "learning_rate": 0.0001, + "loss": 1.0922, + "step": 39244 + }, + { + "epoch": 31.345846645367413, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 39245 + }, + { + "epoch": 31.34664536741214, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0896, + "step": 39246 + }, + { + "epoch": 31.347444089456868, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0893, + "step": 39247 + }, + { + "epoch": 31.3482428115016, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0756, + "step": 39248 + }, + { + "epoch": 31.349041533546327, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0825, + "step": 39249 + }, + { + "epoch": 31.349840255591054, + "grad_norm": 0.21875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 39250 + }, + { + "epoch": 31.350638977635782, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 39251 + }, + { + "epoch": 31.35143769968051, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 39252 + }, + { + "epoch": 31.35223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0864, + "step": 39253 + }, + { + "epoch": 31.35303514376997, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 39254 + }, + { + "epoch": 31.353833865814696, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 39255 + }, + { + "epoch": 31.354632587859424, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 39256 + }, + { + "epoch": 31.355431309904155, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.0956, + "step": 39257 + }, + { + "epoch": 31.356230031948883, + "grad_norm": 0.18359375, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 39258 + }, + { + "epoch": 31.35702875399361, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 39259 + }, + { + "epoch": 31.357827476038338, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0903, + "step": 39260 + }, + { + "epoch": 31.358626198083066, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 39261 + }, + { + "epoch": 31.359424920127797, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 39262 + }, + { + "epoch": 31.360223642172524, + "grad_norm": 0.197265625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 39263 + }, + { + "epoch": 31.361022364217252, + "grad_norm": 0.1328125, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 39264 + }, + { + "epoch": 31.36182108626198, + "grad_norm": 0.166015625, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 39265 + }, + { + "epoch": 31.36261980830671, + "grad_norm": 0.2080078125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 39266 + }, + { + "epoch": 31.36341853035144, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 39267 + }, + { + "epoch": 31.364217252396166, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0931, + "step": 39268 + }, + { + "epoch": 31.365015974440894, + "grad_norm": 0.1513671875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 39269 + }, + { + "epoch": 31.36581469648562, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 39270 + }, + { + "epoch": 31.366613418530353, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 39271 + }, + { + "epoch": 31.36741214057508, + "grad_norm": 0.11279296875, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 39272 + }, + { + "epoch": 31.368210862619808, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 39273 + }, + { + "epoch": 31.369009584664536, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0856, + "step": 39274 + }, + { + "epoch": 31.369808306709267, + "grad_norm": 0.107421875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 39275 + }, + { + "epoch": 31.370607028753994, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 39276 + }, + { + "epoch": 31.371405750798722, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0944, + "step": 39277 + }, + { + "epoch": 31.37220447284345, + "grad_norm": 0.11328125, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 39278 + }, + { + "epoch": 31.373003194888177, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 39279 + }, + { + "epoch": 31.37380191693291, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0823, + "step": 39280 + }, + { + "epoch": 31.374600638977636, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0884, + "step": 39281 + }, + { + "epoch": 31.375399361022364, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 39282 + }, + { + "epoch": 31.37619808306709, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 39283 + }, + { + "epoch": 31.376996805111823, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 39284 + }, + { + "epoch": 31.37779552715655, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 39285 + }, + { + "epoch": 31.378594249201278, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 39286 + }, + { + "epoch": 31.379392971246006, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 39287 + }, + { + "epoch": 31.380191693290733, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 39288 + }, + { + "epoch": 31.380990415335464, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0918, + "step": 39289 + }, + { + "epoch": 31.381789137380192, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0914, + "step": 39290 + }, + { + "epoch": 31.38258785942492, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 39291 + }, + { + "epoch": 31.383386581469647, + "grad_norm": 0.1142578125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 39292 + }, + { + "epoch": 31.38418530351438, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 39293 + }, + { + "epoch": 31.384984025559106, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 39294 + }, + { + "epoch": 31.385782747603834, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 39295 + }, + { + "epoch": 31.38658146964856, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 39296 + }, + { + "epoch": 31.38738019169329, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0899, + "step": 39297 + }, + { + "epoch": 31.38817891373802, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 39298 + }, + { + "epoch": 31.388977635782748, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0814, + "step": 39299 + }, + { + "epoch": 31.389776357827476, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0776, + "step": 39300 + }, + { + "epoch": 31.390575079872203, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0976, + "step": 39301 + }, + { + "epoch": 31.391373801916934, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 39302 + }, + { + "epoch": 31.392172523961662, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0836, + "step": 39303 + }, + { + "epoch": 31.39297124600639, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 39304 + }, + { + "epoch": 31.393769968051117, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 39305 + }, + { + "epoch": 31.394568690095845, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 39306 + }, + { + "epoch": 31.395367412140576, + "grad_norm": 0.1875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 39307 + }, + { + "epoch": 31.396166134185304, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 39308 + }, + { + "epoch": 31.39696485623003, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 39309 + }, + { + "epoch": 31.39776357827476, + "grad_norm": 0.265625, + "learning_rate": 0.0001, + "loss": 1.0901, + "step": 39310 + }, + { + "epoch": 31.39856230031949, + "grad_norm": 0.1201171875, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 39311 + }, + { + "epoch": 31.399361022364218, + "grad_norm": 0.1748046875, + "learning_rate": 0.0001, + "loss": 1.0866, + "step": 39312 + }, + { + "epoch": 31.400159744408946, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0795, + "step": 39313 + }, + { + "epoch": 31.400958466453673, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 39314 + }, + { + "epoch": 31.4017571884984, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0838, + "step": 39315 + }, + { + "epoch": 31.402555910543132, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.083, + "step": 39316 + }, + { + "epoch": 31.40335463258786, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 39317 + }, + { + "epoch": 31.404153354632587, + "grad_norm": 0.10986328125, + "learning_rate": 0.0001, + "loss": 1.0808, + "step": 39318 + }, + { + "epoch": 31.404952076677315, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 39319 + }, + { + "epoch": 31.405750798722046, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0806, + "step": 39320 + }, + { + "epoch": 31.406549520766774, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0881, + "step": 39321 + }, + { + "epoch": 31.4073482428115, + "grad_norm": 0.125, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 39322 + }, + { + "epoch": 31.40814696485623, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 39323 + }, + { + "epoch": 31.408945686900957, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 39324 + }, + { + "epoch": 31.409744408945688, + "grad_norm": 0.1318359375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 39325 + }, + { + "epoch": 31.410543130990416, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 39326 + }, + { + "epoch": 31.411341853035143, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 39327 + }, + { + "epoch": 31.41214057507987, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0826, + "step": 39328 + }, + { + "epoch": 31.412939297124602, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0807, + "step": 39329 + }, + { + "epoch": 31.41373801916933, + "grad_norm": 0.12060546875, + "learning_rate": 0.0001, + "loss": 1.0924, + "step": 39330 + }, + { + "epoch": 31.414536741214057, + "grad_norm": 0.2041015625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 39331 + }, + { + "epoch": 31.415335463258785, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 39332 + }, + { + "epoch": 31.416134185303516, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 39333 + }, + { + "epoch": 31.416932907348244, + "grad_norm": 0.13671875, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 39334 + }, + { + "epoch": 31.41773162939297, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0766, + "step": 39335 + }, + { + "epoch": 31.4185303514377, + "grad_norm": 0.1376953125, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 39336 + }, + { + "epoch": 31.419329073482427, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0952, + "step": 39337 + }, + { + "epoch": 31.420127795527158, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 39338 + }, + { + "epoch": 31.420926517571885, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 39339 + }, + { + "epoch": 31.421725239616613, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 39340 + }, + { + "epoch": 31.42252396166134, + "grad_norm": 0.19140625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 39341 + }, + { + "epoch": 31.423322683706072, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 39342 + }, + { + "epoch": 31.4241214057508, + "grad_norm": 0.193359375, + "learning_rate": 0.0001, + "loss": 1.0962, + "step": 39343 + }, + { + "epoch": 31.424920127795527, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 39344 + }, + { + "epoch": 31.425718849840255, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0904, + "step": 39345 + }, + { + "epoch": 31.426517571884983, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 39346 + }, + { + "epoch": 31.427316293929714, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 39347 + }, + { + "epoch": 31.42811501597444, + "grad_norm": 0.2109375, + "learning_rate": 0.0001, + "loss": 1.0837, + "step": 39348 + }, + { + "epoch": 31.42891373801917, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.1003, + "step": 39349 + }, + { + "epoch": 31.429712460063897, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 39350 + }, + { + "epoch": 31.430511182108628, + "grad_norm": 0.140625, + "learning_rate": 0.0001, + "loss": 1.0744, + "step": 39351 + }, + { + "epoch": 31.431309904153355, + "grad_norm": 0.126953125, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 39352 + }, + { + "epoch": 31.432108626198083, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.091, + "step": 39353 + }, + { + "epoch": 31.43290734824281, + "grad_norm": 0.2314453125, + "learning_rate": 0.0001, + "loss": 1.0862, + "step": 39354 + }, + { + "epoch": 31.43370607028754, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0909, + "step": 39355 + }, + { + "epoch": 31.43450479233227, + "grad_norm": 0.29296875, + "learning_rate": 0.0001, + "loss": 1.0882, + "step": 39356 + }, + { + "epoch": 31.435303514376997, + "grad_norm": 0.1552734375, + "learning_rate": 0.0001, + "loss": 1.092, + "step": 39357 + }, + { + "epoch": 31.436102236421725, + "grad_norm": 0.2490234375, + "learning_rate": 0.0001, + "loss": 1.0929, + "step": 39358 + }, + { + "epoch": 31.436900958466452, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 39359 + }, + { + "epoch": 31.437699680511184, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0762, + "step": 39360 + }, + { + "epoch": 31.43849840255591, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0774, + "step": 39361 + }, + { + "epoch": 31.43929712460064, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0822, + "step": 39362 + }, + { + "epoch": 31.440095846645367, + "grad_norm": 0.2412109375, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 39363 + }, + { + "epoch": 31.440894568690094, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 39364 + }, + { + "epoch": 31.441693290734825, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0907, + "step": 39365 + }, + { + "epoch": 31.442492012779553, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 39366 + }, + { + "epoch": 31.44329073482428, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0831, + "step": 39367 + }, + { + "epoch": 31.44408945686901, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.081, + "step": 39368 + }, + { + "epoch": 31.44488817891374, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 39369 + }, + { + "epoch": 31.445686900958467, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0775, + "step": 39370 + }, + { + "epoch": 31.446485623003195, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 39371 + }, + { + "epoch": 31.447284345047922, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0887, + "step": 39372 + }, + { + "epoch": 31.44808306709265, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.086, + "step": 39373 + }, + { + "epoch": 31.44888178913738, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0915, + "step": 39374 + }, + { + "epoch": 31.44968051118211, + "grad_norm": 0.1923828125, + "learning_rate": 0.0001, + "loss": 1.0894, + "step": 39375 + }, + { + "epoch": 31.450479233226837, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 39376 + }, + { + "epoch": 31.451277955271564, + "grad_norm": 0.2431640625, + "learning_rate": 0.0001, + "loss": 1.1043, + "step": 39377 + }, + { + "epoch": 31.452076677316295, + "grad_norm": 0.203125, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 39378 + }, + { + "epoch": 31.452875399361023, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 39379 + }, + { + "epoch": 31.45367412140575, + "grad_norm": 0.23046875, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 39380 + }, + { + "epoch": 31.45447284345048, + "grad_norm": 0.228515625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 39381 + }, + { + "epoch": 31.455271565495206, + "grad_norm": 0.224609375, + "learning_rate": 0.0001, + "loss": 1.0772, + "step": 39382 + }, + { + "epoch": 31.456070287539937, + "grad_norm": 0.2001953125, + "learning_rate": 0.0001, + "loss": 1.0939, + "step": 39383 + }, + { + "epoch": 31.456869009584665, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.079, + "step": 39384 + }, + { + "epoch": 31.457667731629392, + "grad_norm": 0.1708984375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 39385 + }, + { + "epoch": 31.45846645367412, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0897, + "step": 39386 + }, + { + "epoch": 31.45926517571885, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0955, + "step": 39387 + }, + { + "epoch": 31.46006389776358, + "grad_norm": 0.1396484375, + "learning_rate": 0.0001, + "loss": 1.0816, + "step": 39388 + }, + { + "epoch": 31.460862619808307, + "grad_norm": 0.1767578125, + "learning_rate": 0.0001, + "loss": 1.0834, + "step": 39389 + }, + { + "epoch": 31.461661341853034, + "grad_norm": 0.1279296875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 39390 + }, + { + "epoch": 31.462460063897762, + "grad_norm": 0.12109375, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 39391 + }, + { + "epoch": 31.463258785942493, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0906, + "step": 39392 + }, + { + "epoch": 31.46405750798722, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0908, + "step": 39393 + }, + { + "epoch": 31.46485623003195, + "grad_norm": 0.12255859375, + "learning_rate": 0.0001, + "loss": 1.0889, + "step": 39394 + }, + { + "epoch": 31.465654952076676, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0839, + "step": 39395 + }, + { + "epoch": 31.466453674121407, + "grad_norm": 0.1064453125, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 39396 + }, + { + "epoch": 31.467252396166135, + "grad_norm": 0.1181640625, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 39397 + }, + { + "epoch": 31.468051118210862, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 39398 + }, + { + "epoch": 31.46884984025559, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0892, + "step": 39399 + }, + { + "epoch": 31.46964856230032, + "grad_norm": 0.1630859375, + "learning_rate": 0.0001, + "loss": 1.0977, + "step": 39400 + }, + { + "epoch": 31.47044728434505, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 39401 + }, + { + "epoch": 31.471246006389777, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 39402 + }, + { + "epoch": 31.472044728434504, + "grad_norm": 0.26953125, + "learning_rate": 0.0001, + "loss": 1.0937, + "step": 39403 + }, + { + "epoch": 31.472843450479232, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 39404 + }, + { + "epoch": 31.473642172523963, + "grad_norm": 0.2021484375, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 39405 + }, + { + "epoch": 31.47444089456869, + "grad_norm": 0.275390625, + "learning_rate": 0.0001, + "loss": 1.0936, + "step": 39406 + }, + { + "epoch": 31.47523961661342, + "grad_norm": 0.1689453125, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 39407 + }, + { + "epoch": 31.476038338658146, + "grad_norm": 0.2578125, + "learning_rate": 0.0001, + "loss": 1.0875, + "step": 39408 + }, + { + "epoch": 31.476837060702877, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0846, + "step": 39409 + }, + { + "epoch": 31.477635782747605, + "grad_norm": 0.220703125, + "learning_rate": 0.0001, + "loss": 1.0972, + "step": 39410 + }, + { + "epoch": 31.478434504792332, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 39411 + }, + { + "epoch": 31.47923322683706, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0923, + "step": 39412 + }, + { + "epoch": 31.480031948881788, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0793, + "step": 39413 + }, + { + "epoch": 31.48083067092652, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 39414 + }, + { + "epoch": 31.481629392971247, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0932, + "step": 39415 + }, + { + "epoch": 31.482428115015974, + "grad_norm": 0.11962890625, + "learning_rate": 0.0001, + "loss": 1.0867, + "step": 39416 + }, + { + "epoch": 31.483226837060702, + "grad_norm": 0.1953125, + "learning_rate": 0.0001, + "loss": 1.0817, + "step": 39417 + }, + { + "epoch": 31.484025559105433, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0951, + "step": 39418 + }, + { + "epoch": 31.48482428115016, + "grad_norm": 0.16796875, + "learning_rate": 0.0001, + "loss": 1.0925, + "step": 39419 + }, + { + "epoch": 31.48562300319489, + "grad_norm": 0.1611328125, + "learning_rate": 0.0001, + "loss": 1.0843, + "step": 39420 + }, + { + "epoch": 31.486421725239616, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 39421 + }, + { + "epoch": 31.487220447284344, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 39422 + }, + { + "epoch": 31.488019169329075, + "grad_norm": 0.1962890625, + "learning_rate": 0.0001, + "loss": 1.084, + "step": 39423 + }, + { + "epoch": 31.488817891373802, + "grad_norm": 0.1728515625, + "learning_rate": 0.0001, + "loss": 1.0784, + "step": 39424 + }, + { + "epoch": 31.48961661341853, + "grad_norm": 0.189453125, + "learning_rate": 0.0001, + "loss": 1.0865, + "step": 39425 + }, + { + "epoch": 31.490415335463258, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0803, + "step": 39426 + }, + { + "epoch": 31.49121405750799, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0917, + "step": 39427 + }, + { + "epoch": 31.492012779552716, + "grad_norm": 0.1845703125, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 39428 + }, + { + "epoch": 31.492811501597444, + "grad_norm": 0.1572265625, + "learning_rate": 0.0001, + "loss": 1.0852, + "step": 39429 + }, + { + "epoch": 31.49361022364217, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.0849, + "step": 39430 + }, + { + "epoch": 31.4944089456869, + "grad_norm": 0.201171875, + "learning_rate": 0.0001, + "loss": 1.0873, + "step": 39431 + }, + { + "epoch": 31.49520766773163, + "grad_norm": 0.119140625, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 39432 + }, + { + "epoch": 31.49600638977636, + "grad_norm": 0.2451171875, + "learning_rate": 0.0001, + "loss": 1.0746, + "step": 39433 + }, + { + "epoch": 31.496805111821086, + "grad_norm": 0.173828125, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 39434 + }, + { + "epoch": 31.497603833865814, + "grad_norm": 0.154296875, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 39435 + }, + { + "epoch": 31.498402555910545, + "grad_norm": 0.1982421875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 39436 + }, + { + "epoch": 31.499201277955272, + "grad_norm": 0.158203125, + "learning_rate": 0.0001, + "loss": 1.1005, + "step": 39437 + }, + { + "epoch": 31.5, + "grad_norm": 0.244140625, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 39438 + }, + { + "epoch": 31.500798722044728, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0827, + "step": 39439 + }, + { + "epoch": 31.501597444089455, + "grad_norm": 0.169921875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 39440 + }, + { + "epoch": 31.502396166134186, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.0857, + "step": 39441 + }, + { + "epoch": 31.503194888178914, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 39442 + }, + { + "epoch": 31.50399361022364, + "grad_norm": 0.1796875, + "learning_rate": 0.0001, + "loss": 1.0958, + "step": 39443 + }, + { + "epoch": 31.50479233226837, + "grad_norm": 0.1416015625, + "learning_rate": 0.0001, + "loss": 1.0989, + "step": 39444 + }, + { + "epoch": 31.5055910543131, + "grad_norm": 0.216796875, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 39445 + }, + { + "epoch": 31.50638977635783, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 39446 + }, + { + "epoch": 31.507188498402556, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0961, + "step": 39447 + }, + { + "epoch": 31.507987220447284, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.09, + "step": 39448 + }, + { + "epoch": 31.50878594249201, + "grad_norm": 0.2099609375, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 39449 + }, + { + "epoch": 31.509584664536742, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 39450 + }, + { + "epoch": 31.51038338658147, + "grad_norm": 0.1826171875, + "learning_rate": 0.0001, + "loss": 1.089, + "step": 39451 + }, + { + "epoch": 31.511182108626198, + "grad_norm": 0.1455078125, + "learning_rate": 0.0001, + "loss": 1.0869, + "step": 39452 + }, + { + "epoch": 31.511980830670925, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 39453 + }, + { + "epoch": 31.512779552715656, + "grad_norm": 0.1123046875, + "learning_rate": 0.0001, + "loss": 1.0833, + "step": 39454 + }, + { + "epoch": 31.513578274760384, + "grad_norm": 0.11767578125, + "learning_rate": 0.0001, + "loss": 1.0916, + "step": 39455 + }, + { + "epoch": 31.51437699680511, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0863, + "step": 39456 + }, + { + "epoch": 31.51517571884984, + "grad_norm": 0.1337890625, + "learning_rate": 0.0001, + "loss": 1.0912, + "step": 39457 + }, + { + "epoch": 31.515974440894567, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0948, + "step": 39458 + }, + { + "epoch": 31.516773162939298, + "grad_norm": 0.10888671875, + "learning_rate": 0.0001, + "loss": 1.099, + "step": 39459 + }, + { + "epoch": 31.517571884984026, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0835, + "step": 39460 + }, + { + "epoch": 31.518370607028753, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0891, + "step": 39461 + }, + { + "epoch": 31.51916932907348, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0855, + "step": 39462 + }, + { + "epoch": 31.519968051118212, + "grad_norm": 0.171875, + "learning_rate": 0.0001, + "loss": 1.0769, + "step": 39463 + }, + { + "epoch": 31.52076677316294, + "grad_norm": 0.123046875, + "learning_rate": 0.0001, + "loss": 1.0832, + "step": 39464 + }, + { + "epoch": 31.521565495207668, + "grad_norm": 0.130859375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 39465 + }, + { + "epoch": 31.522364217252395, + "grad_norm": 0.11669921875, + "learning_rate": 0.0001, + "loss": 1.0858, + "step": 39466 + }, + { + "epoch": 31.523162939297123, + "grad_norm": 0.16015625, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 39467 + }, + { + "epoch": 31.523961661341854, + "grad_norm": 0.150390625, + "learning_rate": 0.0001, + "loss": 1.0797, + "step": 39468 + }, + { + "epoch": 31.52476038338658, + "grad_norm": 0.11474609375, + "learning_rate": 0.0001, + "loss": 1.078, + "step": 39469 + }, + { + "epoch": 31.52555910543131, + "grad_norm": 0.134765625, + "learning_rate": 0.0001, + "loss": 1.0841, + "step": 39470 + }, + { + "epoch": 31.526357827476037, + "grad_norm": 0.2177734375, + "learning_rate": 0.0001, + "loss": 1.0868, + "step": 39471 + }, + { + "epoch": 31.527156549520768, + "grad_norm": 0.22265625, + "learning_rate": 0.0001, + "loss": 1.095, + "step": 39472 + }, + { + "epoch": 31.527955271565496, + "grad_norm": 0.142578125, + "learning_rate": 0.0001, + "loss": 1.0927, + "step": 39473 + }, + { + "epoch": 31.528753993610223, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0844, + "step": 39474 + }, + { + "epoch": 31.52955271565495, + "grad_norm": 0.15234375, + "learning_rate": 0.0001, + "loss": 1.0895, + "step": 39475 + }, + { + "epoch": 31.53035143769968, + "grad_norm": 0.1298828125, + "learning_rate": 0.0001, + "loss": 1.0926, + "step": 39476 + }, + { + "epoch": 31.53115015974441, + "grad_norm": 0.1650390625, + "learning_rate": 0.0001, + "loss": 1.0761, + "step": 39477 + }, + { + "epoch": 31.531948881789138, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 39478 + }, + { + "epoch": 31.532747603833865, + "grad_norm": 0.1259765625, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 39479 + }, + { + "epoch": 31.533546325878593, + "grad_norm": 0.1787109375, + "learning_rate": 0.0001, + "loss": 1.0876, + "step": 39480 + }, + { + "epoch": 31.534345047923324, + "grad_norm": 0.162109375, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 39481 + }, + { + "epoch": 31.53514376996805, + "grad_norm": 0.12890625, + "learning_rate": 0.0001, + "loss": 1.0861, + "step": 39482 + }, + { + "epoch": 31.53594249201278, + "grad_norm": 0.1435546875, + "learning_rate": 0.0001, + "loss": 1.0824, + "step": 39483 + }, + { + "epoch": 31.536741214057507, + "grad_norm": 0.1494140625, + "learning_rate": 0.0001, + "loss": 1.0871, + "step": 39484 + }, + { + "epoch": 31.537539936102235, + "grad_norm": 0.1171875, + "learning_rate": 0.0001, + "loss": 1.0938, + "step": 39485 + }, + { + "epoch": 31.538338658146966, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.082, + "step": 39486 + }, + { + "epoch": 31.539137380191693, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0886, + "step": 39487 + }, + { + "epoch": 31.53993610223642, + "grad_norm": 0.1015625, + "learning_rate": 0.0001, + "loss": 1.0829, + "step": 39488 + }, + { + "epoch": 31.54073482428115, + "grad_norm": 0.1357421875, + "learning_rate": 0.0001, + "loss": 1.0872, + "step": 39489 + }, + { + "epoch": 31.54153354632588, + "grad_norm": 0.1533203125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 39490 + }, + { + "epoch": 31.542332268370608, + "grad_norm": 0.138671875, + "learning_rate": 0.0001, + "loss": 1.0859, + "step": 39491 + }, + { + "epoch": 31.543130990415335, + "grad_norm": 0.12451171875, + "learning_rate": 0.0001, + "loss": 1.0815, + "step": 39492 + }, + { + "epoch": 31.543929712460063, + "grad_norm": 0.17578125, + "learning_rate": 0.0001, + "loss": 1.0885, + "step": 39493 + }, + { + "epoch": 31.544728434504794, + "grad_norm": 0.15625, + "learning_rate": 0.0001, + "loss": 1.0921, + "step": 39494 + }, + { + "epoch": 31.54552715654952, + "grad_norm": 0.185546875, + "learning_rate": 0.0001, + "loss": 1.0878, + "step": 39495 + }, + { + "epoch": 31.54632587859425, + "grad_norm": 0.146484375, + "learning_rate": 0.0001, + "loss": 1.0854, + "step": 39496 + }, + { + "epoch": 31.547124600638977, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0771, + "step": 39497 + }, + { + "epoch": 31.547923322683705, + "grad_norm": 0.14453125, + "learning_rate": 0.0001, + "loss": 1.0851, + "step": 39498 + }, + { + "epoch": 31.548722044728436, + "grad_norm": 0.1474609375, + "learning_rate": 0.0001, + "loss": 1.0805, + "step": 39499 + }, + { + "epoch": 31.549520766773163, + "grad_norm": 0.1484375, + "learning_rate": 0.0001, + "loss": 1.0879, + "step": 39500 + } + ], + "logging_steps": 1.0, + "max_steps": 80128, + "num_input_tokens_seen": 0, + "num_train_epochs": 64, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.2611631492892721e+19, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +}